Issuu on Google+

Book of Vaadin Vaadin 6.4


Book of Vaadin: Vaadin 6.4 Vaadin Ltd Marko GrÜnroos Vaadin Framework 6.4.8 Published: 2010-11-08 Copyright Š 2000-2010 Vaadin Ltd

Abstract Vaadin is a server-side AJAX web application development framework that enables developers to build high-quality user interfaces with Java. It provides a library of ready-to-use user interface components and a clean framework for creating your own components. The focus is on ease-of-use, re-usability, extensibility, and meeting the requirements of large enterprise applications. Vaadin has been used in production since 2001 and it has proven to be suitable for building demanding business applications. All rights reserved.


Table of Contents Preface ............................................................................................................................ ix 1. Introduction ................................................................................................................. 1.1. Overview ............................................................................................................ 1.2. Example Application Walkthrough ........................................................................ 1.3. Support for the Eclipse IDE ................................................................................. 1.4. Goals and Philosophy ......................................................................................... 1.5. Background ........................................................................................................

1 1 3 4 4 5

2. Getting Started with Vaadin ......................................................................................... 7 2.1. Installing Vaadin ................................................................................................. 7 2.1.1. Installing the Distribution Package ............................................................. 8 2.1.2. Starting the Content Browser .................................................................... 8 2.1.3. Package Contents .................................................................................... 9 2.1.4. Demo Applications ................................................................................. 11 2.2. Setting up the Development Environment ........................................................... 11 2.2.1. Installing Java SDK ................................................................................ 12 2.2.2. Installing Eclipse IDE .............................................................................. 13 2.2.3. Installing Apache Tomcat ........................................................................ 14 2.2.4. Firefox and Firebug ................................................................................ 14 2.2.5. Vaadin Plugin for Eclipse ........................................................................ 15 2.3. QuickStart with Eclipse ...................................................................................... 18 2.3.1. Starting Eclipse ...................................................................................... 18 2.3.2. Importing Vaadin as a Project ................................................................. 18 2.3.3. Running the Demo Applications in Eclipse ............................................... 19 2.3.4. Debugging the Demo Applications in Eclipse ........................................... 23 2.3.5. Using QuickStart as a Project Skeleton ................................................... 23 2.4. Your First Project with Vaadin ............................................................................. 24 2.4.1. Creating the Project ............................................................................... 24 2.4.2. Exploring the Project .............................................................................. 28 2.4.3. Setting Up and Starting the Web Server .................................................. 30 2.4.4. Running and Debugging ......................................................................... 32 3. Architecture ............................................................................................................... 3.1. Overview .......................................................................................................... 3.2. Technological Background ................................................................................. 3.2.1. AJAX ..................................................................................................... 3.2.2. Google Web Toolkit ................................................................................ 3.2.3. JSON .................................................................................................... 3.3. Applications as Java Servlet Sessions ................................................................ 3.4. Client-Side Engine ............................................................................................ 3.5. Events and Listeners .........................................................................................

35 35 38 38 38 39 39 39 41

4. Writing a Web Application .......................................................................................... 4.1. Overview .......................................................................................................... 4.2. Managing the Main Window ............................................................................... 4.3. Child Windows .................................................................................................. 4.3.1. Opening and Closing a Child Window ...................................................... 4.3.2. Window Positioning ................................................................................ 4.3.3. Modal Windows ...................................................................................... 4.4. Handling Events with Listeners .......................................................................... 4.5. Referencing Resources ..................................................................................... 4.5.1. Resource Interfaces and Classes ............................................................

43 43 46 46 47 49 49 50 52 53

Book of Vaadin

iii


Book of Vaadin

4.5.2. File Resources ....................................................................................... 53 4.5.3. Class Loader Resources ........................................................................ 54 4.5.4. Theme Resources .................................................................................. 54 4.5.5. Stream Resources ................................................................................. 54 4.6. Shutting Down an Application ............................................................................ 56 4.6.1. Closing an Application ............................................................................ 56 4.6.2. Handling the Closing of a Window ........................................................... 56 4.7. Handling Errors ................................................................................................. 57 4.7.1. Error Indicator and message ................................................................... 57 4.7.2. Notifications ........................................................................................... 57 4.7.3. Handling Uncaught Exceptions .............................................................. 60 4.8. Setting Up the Application Environment .............................................................. 61 4.8.1. Creating Deployable WAR in Eclipse ....................................................... 61 4.8.2. Web Application Contents ....................................................................... 61 4.8.3. Deployment Descriptor web.xml ............................................................ 62 5. User Interface Components ....................................................................................... 65 5.1. Overview .......................................................................................................... 66 5.2. Interfaces and Abstractions ............................................................................... 67 5.2.1. Component Interface ............................................................................. 68 5.2.2. AbstractComponent ............................................................................. 69 5.2.3. Field Components (Field and AbstractField) .......................................... 69 5.3. Common Component Features .......................................................................... 72 5.3.1. Caption ................................................................................................. 72 5.3.2. Description and Tooltips .......................................................................... 73 5.3.3. Enabled ................................................................................................. 74 5.3.4. Icon ....................................................................................................... 75 5.3.5. Locale ................................................................................................... 75 5.3.6. Read-Only ............................................................................................. 78 5.3.7. Style Name ............................................................................................ 78 5.3.8. Visible ................................................................................................... 79 5.3.9. Sizing Components ................................................................................ 80 5.3.10. Managing Input Focus .......................................................................... 81 5.4. Label ............................................................................................................... 82 5.5. Link ................................................................................................................. 85 5.6. TextField ......................................................................................................... 86 5.7. RichTextArea .................................................................................................. 87 5.8. Date and Time Input with DateField ................................................................... 88 5.8.1. PopupDateField .................................................................................... 89 5.8.2. InlineDateField ..................................................................................... 92 5.8.3. Time Resolution ..................................................................................... 93 5.8.4. DateField Locale .................................................................................... 93 5.9. Button ............................................................................................................. 93 5.10. CheckBox ...................................................................................................... 94 5.11. Selecting Items ............................................................................................... 95 5.11.1. Binding Selection Components to Data .................................................. 95 5.11.2. Basic Select Component ...................................................................... 99 5.11.3. ListSelect ......................................................................................... 101 5.11.4. Native Selection Component NativeSelect .......................................... 102 5.11.5. Radio Button and Check Box Groups with OptionGroup ...................... 102 5.11.6. Twin Column Selection with TwinColSelect ......................................... 104 5.11.7. Allowing Adding New Items ................................................................. 104 5.11.8. Multiple Selection Mode ...................................................................... 105 5.11.9. Other Common Features .................................................................... 106

iv


Book of Vaadin

5.12. Table ............................................................................................................ 5.12.1. Selecting Items in a Table ................................................................... 5.12.2. Table Features ................................................................................... 5.12.3. Editing the Values in a Table ................................................................ 5.12.4. Column Headers and Footers .............................................................. 5.12.5. Generated Table Columns ................................................................... 5.12.6. Formatting Table Columns ................................................................... 5.12.7. CSS Style Rules ................................................................................. 5.13. Tree ............................................................................................................. 5.14. MenuBar ...................................................................................................... 5.15. Embedded ................................................................................................... 5.15.1. Embedded Objects ............................................................................. 5.15.2. Embedded Images ............................................................................. 5.15.3. Browser Frames ................................................................................. 5.16. Upload ......................................................................................................... 5.17. Form ............................................................................................................ 5.17.1. Form as a User Interface Component .................................................. 5.17.2. Binding Form to Data .......................................................................... 5.17.3. Validating Form Input .......................................................................... 5.17.4. Buffering Form Data ........................................................................... 5.18. ProgressIndicator ....................................................................................... 5.18.1. Doing Heavy Computation .................................................................. 5.19. Slider ........................................................................................................... 5.20. LoginForm ................................................................................................... 5.20.1. Customizing LoginForm .................................................................... 5.21. Component Composition with CustomComponent .........................................

106 108 109 112 115 117 120 121 123 125 126 127 127 128 128 131 131 133 136 138 139 139 140 142 143 145

6. Managing Layout ..................................................................................................... 6.1. Overview ........................................................................................................ 6.2. Window and Panel Root Layout ........................................................................ 6.3. VerticalLayout and HorizontalLayout ............................................................. 6.3.1. Sizing Contained Components .............................................................. 6.4. GridLayout .................................................................................................... 6.4.1. Sizing Grid Cells .................................................................................. 6.5. FormLayout ................................................................................................... 6.6. Panel ............................................................................................................. 6.7. SplitPanel ...................................................................................................... 6.8. TabSheet ....................................................................................................... 6.9. Accordion ..................................................................................................... 6.10. AbsoluteLayout ........................................................................................... 6.11. CssLayout ................................................................................................... 6.12. Layout Formatting ......................................................................................... 6.12.1. Layout Size ........................................................................................ 6.12.2. Layout Cell Alignment ......................................................................... 6.12.3. Layout Cell Spacing ............................................................................ 6.12.4. Layout Margins ................................................................................... 6.13. Custom Layouts ............................................................................................

147 148 150 150 151 154 155 158 159 161 163 166 167 170 172 172 174 175 177 178

7. Visual User Interface Design with Eclipse (experimental) ........................................ 7.1. Overview ........................................................................................................ 7.2. Creating a New CustomComponent ............................................................... 7.3. Using The Visual Editor ................................................................................... 7.3.1. Adding New Components ..................................................................... 7.3.2. Setting Component Properties ..............................................................

181 182 182 184 184 185

v


Book of Vaadin

vi

7.3.3. Editing an AbsoluteLayout .................................................................. 7.4. Structure of a Visually Editable Component ...................................................... 7.4.1. Sub-Component References ................................................................. 7.4.2. Sub-Component Builders ...................................................................... 7.4.3. The Constructor ...................................................................................

188 189 190 190 191

8. Themes .................................................................................................................... 8.1. Overview ........................................................................................................ 8.2. Introduction to Cascading Style Sheets ............................................................ 8.2.1. Basic CSS Rules .................................................................................. 8.2.2. Matching by Element Class ................................................................... 8.2.3. Matching by Descendant Relationship ................................................... 8.2.4. Notes on Compatibility .......................................................................... 8.3. Creating and Using Themes ............................................................................ 8.3.1. Styling Standard Components ............................................................... 8.3.2. Built-in Themes .................................................................................... 8.3.3. Using Themes ...................................................................................... 8.3.4. Theme Inheritance ............................................................................... 8.4. Creating a Theme in Eclipse ............................................................................

193 193 195 195 196 197 199 200 200 202 203 203 204

9. Binding Components to Data ................................................................................... 9.1. Overview ........................................................................................................ 9.2. Properties ...................................................................................................... 9.2.1. Property Viewers and Editors ................................................................ 9.2.2. ObjectProperty Implementation ........................................................... 9.2.3. Implementing the Property Interface ..................................................... 9.3. Holding properties in Items .............................................................................. 9.3.1. The PropertysetItem Implementation ................................................... 9.3.2. Wrapping a Bean in a BeanItem ........................................................... 9.4. Collecting items in Containers .......................................................................... 9.4.1. Iterating Over a Container .....................................................................

207 207 209 210 211 211 212 213 213 214 214

10. Developing Custom Components .......................................................................... 10.1. Overview ...................................................................................................... 10.2. Doing It the Simple Way in Eclipse ................................................................. 10.2.1. Creating a Widget ............................................................................... 10.2.2. Recompiling the Widget Set ................................................................ 10.2.3. Plugin Related Project Settings ........................................................... 10.3. Google Web Toolkit Widgets ........................................................................... 10.3.1. Extending a Vaadin Widget ................................................................. 10.3.2. Example: A Color Picker GWT Widget ................................................. 10.3.3. Styling GWT Widgets .......................................................................... 10.4. Integrating a GWT Widget .............................................................................. 10.4.1. Deserialization of Component State from Server .................................. 10.4.2. Serialization of Component State to Server .......................................... 10.4.3. Example: Integrating the Color Picker Widget ....................................... 10.5. Defining a Widget Set .................................................................................... 10.6. Server-Side Components ............................................................................... 10.6.1. Binding to the Client-Side Widget ........................................................ 10.6.2. Server-Client Serialization ................................................................... 10.6.3. Client-Server Deserialization ............................................................... 10.6.4. Example: Color Picker Server-Side Component .................................... 10.7. Using a Custom Component .......................................................................... 10.7.1. Example: Color Picker Application ....................................................... 10.7.2. Web Application Deployment ...............................................................

217 218 220 220 222 223 224 225 226 228 229 230 231 232 234 235 235 235 235 236 237 237 238


Book of Vaadin

10.8. GWT Widget Development ............................................................................. 10.8.1. Creating a Widget Project ................................................................... 10.8.2. Importing GWT Installation Package .................................................... 10.8.3. Writing the Code ................................................................................ 10.8.4. Compiling GWT Widget Sets ............................................................... 10.8.5. Ready to Run ..................................................................................... 10.8.6. GWT Development Mode .................................................................... 10.8.7. Packaging a Widget Set ...................................................................... 10.8.8. Troubleshooting ..................................................................................

239 239 240 241 241 245 246 251 251

11. Advanced Web Application Topics ......................................................................... 11.1. Special Characteristics of AJAX Applications .................................................. 11.2. Application-Level Windows ............................................................................ 11.2.1. Creating New Application-Level Windows ............................................. 11.2.2. Creating Windows Dynamically ........................................................... 11.2.3. Closing Windows ................................................................................ 11.2.4. Caveats in Using Multiple Windows ..................................................... 11.3. Embedding Applications in Web Pages ........................................................... 11.3.1. Embedding Inside a div Element ........................................................ 11.3.2. Embedding Inside an iframe Element ................................................ 11.4. Debug and Production Mode .......................................................................... 11.4.1. Debug Mode ...................................................................................... 11.4.2. Analyzing Layouts .............................................................................. 11.4.3. Custom Layouts ................................................................................. 11.4.4. Debug Functions for Component Developers ........................................ 11.5. Resources .................................................................................................... 11.5.1. URI Handlers ..................................................................................... 11.5.2. Parameter Handlers ............................................................................ 11.6. Shortcut Keys ............................................................................................... 11.6.1. Click Shortcuts for Default Buttons ....................................................... 11.6.2. Field Focus Shortcuts ......................................................................... 11.6.3. Generic Shortcut Actions .................................................................... 11.6.4. Supported Key Codes and Modifier Keys ............................................. 11.7. Printing ......................................................................................................... 11.8. Portal Integration ........................................................................................... 11.8.1. Deploying to a Portal .......................................................................... 11.8.2. Creating a Portal Application Project in Eclipse .................................... 11.8.3. Portlet Deployment Descriptors ........................................................... 11.8.4. Portlet Hello World .............................................................................. 11.8.5. Installing Vaadin in Liferay ................................................................... 11.8.6. Handling Portlet Requests .................................................................. 11.8.7. Handling Portlet Mode Changes .......................................................... 11.8.8. Non-Vaadin Portlet Modes .................................................................. 11.9. Google App Engine Integration ...................................................................... 11.10. Common Security Issues ............................................................................. 11.10.1. Sanitizing User Input to Prevent Cross-Site Scripting .......................... 11.11. URI Fragment and History Management with UriFragmentUtility .................. 11.12. Capturing HTTP Requests ........................................................................... 11.12.1. Using Request and Response Objects ............................................... 11.12.2. Managing Cookies ............................................................................ 11.13. Drag and Drop ............................................................................................ 11.13.1. Handling Drops ................................................................................ 11.13.2. Dropping Items On a Tree ................................................................. 11.13.3. Dropping Items On a Table ...............................................................

253 254 254 255 256 259 260 261 262 264 266 266 267 268 268 268 269 269 271 271 272 272 274 275 276 276 277 279 284 284 286 287 289 291 292 292 293 294 295 296 297 298 298 300

vii


Book of Vaadin

11.13.4. Accepting Drops ............................................................................... 11.13.5. Dragging Components ...................................................................... 11.13.6. Dropping on a Component ................................................................ 11.13.7. Dragging Files from Outside the Browser ........................................... 11.14. Using Add-on Components .......................................................................... 11.14.1. Downloading Add-ons from Vaadin Directory ...................................... 11.14.2. Compiling Add-on Widget Sets in Eclipse ........................................... 11.14.3. Compiling Add-on Widget Sets with an Ant Script ............................... 11.14.4. Troubleshooting ................................................................................ 11.14.5. Removing Widget Sets ......................................................................

300 303 304 305 305 305 306 306 307 307

A. User Interface Definition Language (UIDL) .............................................................. 309 A.1. API for Painting Components ........................................................................... 310 A.2. JSON Rendering ............................................................................................ 311 B. Songs of Vaadin ...................................................................................................... 315 Index ............................................................................................................................ 319

viii


Preface This book provides an overview of the Vaadin Framework and covers the most important topics which you might encounter when developing applications with it. A more detailed documentation of individual classes, interfaces, and methods is given in the Java API Reference. You can browse an online version of this book at the Vaadin website http://vaadin.com/. A PDF version is also included in the Vaadin installation package and if you install the Vaadin Plugin for Eclipse, you can browse it in the Eclipse Help. You may find the HTML or the Eclipse Help plugin version more easily searchable than this printed book or the PDF version, but the content is the same. Just like the rest of Vaadin, this book is open source. Writing this manual is ongoing work and this edition represents a snapshot at a time when working towards the release of Vaadin 6.4. Nevertheless, many of the Vaadin 6.4 features, such as the enhanced keyboard support and table component enhancements were still in early phase of development at the time of editing, so this edition does not include them. Also, a revised visual editor plugin for Eclipse is expected around the time of Vaadin 6.4, but the chapter describing the editor is about the current version in early 2010. This edition includes an index, which is not yet complete. Many sections are under work and will be expanded in future.

Who is This Book For? This book is intended for software developers who use, or are considering to use, Vaadin to develop web applications. The book assumes that you have some experience with programming in Java, but if not, it is as easy to begin learning Java with Vaadin as with any other UI framework if not easier. No knowledge of AJAX is needed as it is well hidden from the developer. You may have used some desktop-oriented user interface frameworks for Java, such as AWT, Swing, or SWT. Or a library such as Qt for C++. Such knowledge is useful for understanding the scope of Vaadin, the event-driven programming model, and other common concepts of UI frameworks, but not necessary. If you don't have a web graphics designer at hand, knowing the basics of HTML and CSS can help, so that you can develop presentation themes for your application. A brief introduction to CSS is provided. Knowledge of Google Web Toolkit (GWT) may be useful if you develop or integrate new client-side components.

Organization of This Book The Book of Vaadin gives an introduction to what Vaadin is and how you use it to develop web applications. Chapter 1, Introduction

The chapter gives introduction to the application architecture supported by Vaadin, the core design ideas behind the framework, and some historical background.

Chapter 2, Getting Started with Vaadin

This chapter gives practical instructions for installing Vaadin and the reference toolchain, including the Vaadin Plugin for Eclipse, how to run and debug the

Book of Vaadin

ix


Preface

demos, and how to create your own application project in the Eclipse IDE.

x

Chapter 3, Architecture

This chapter gives an introduction to the architecture of Vaadin and its major technologies, including AJAX, Google Web Toolkit, JSON, and event-driven programming.

Chapter 4, Writing a Web Application

This chapter gives all the practical knowledge required for creating applications with Vaadin, such as window management, application lifecycle, deployment in a servlet container, and handling events, errors, and resources.

Chapter 5, User Interface Components

This chapter essentially gives the reference documentation for all the core user interface components in Vaadin and their most significant features. The text gives examples for using each of the components.

Chapter 6, Managing Layout

This chapter describes the layout components, which are used for managing the layout of the user interface, just like in any desktop application frameworks.

Chapter 7, Visual User Interface Design with Eclipse (experimental)

This chapter gives instructions for using the visual editor for Eclipse, which is included in the Vaadin Plugin for the Eclipse IDE.

Chapter 8, Themes

This chapter gives an introduction to Cascading Style Sheets (CSS) and explains how you can use them to build custom visual themes for your application.

Chapter 9, Binding Components to Data

This chapter gives an overview of the built-in data model of Vaadin, consisting of properties, items, and containers.

Chapter 10, Developing Custom Components

This chapter describes the process of creating new client-side widgets with Google Web Toolkit (GWT) and integrating them with server-side counterparts. The chapter also gives practical instructions for creating widget projects in Eclipse, and using the GWT Development Mode to debug widgets.

Chapter 11, Advanced Web Application Topics

This chapter provides many special topics that are commonly needed in applications, such as opening new browser windows, embedding applications in regular web pages, low-level management of resources, shortcut keys, debugging, etc.

Appendix A, User Interface Definition Language (UIDL)

This chapter gives an outline of the low-level UIDL messaging language, normally hidden from the developer. The chapter includes the description of the serialization API needed for synchronizing the component state between the client-side and server-side components.

Organization of This Book


Preface

Appendix B, Songs of Vaadin

Mythological background of the name Vaadin.

Supplementary Material The Vaadin installation package and websites offer plenty of material that can help you understand what Vaadin is, what you can do with it, and how you can do it. Demo Applications

The installation package of Vaadin includes a number of demo applications that you can run and use with your web browser. The content browser allows you to view the source code of the individual demo applications. You should find especially the Sampler demo a good friend of yours. You can find the demo applications online at http://vaadin.com/.

Address Book Tutorial

The Address Book is a sample application accompanied with a tutorial that gives detailed step-by-step instructions for creating a real-life web application with Vaadin. You can find the tutorial from the product website.

Developer's Website

Vaadin Developer's Site at http://dev.vaadin.com/ provides various online resources, such as the ticket system, a development wiki, source repositories, activity timeline, development milestones, and so on. The wiki provides instructions for developers, especially for those who wish to check-out and compile Vaadin itself from the source repository. The technical articles deal with integration of Vaadin applications with various systems, such as JSP, Maven, Spring, Hibernate, and portals. The wiki also provides answers to Frequently Asked Questions.

Online Documentation

You can read this book online at http://vaadin.com/book. Lots of additional material, including technical HOWTOs, answers to Frequently Asked Questions and other documentation is also available on Vaadin web-site [http://dev.vaadin.com/].

Support Stuck with a problem? No need to lose your hair over it, the Vaadin Framework developer community and the Vaadin company offer support for all of your needs. Community Support Forum

You can find the user and developer community forum for Vaadin at http://vaadin.com/forum. Please use the forum to discuss any problems you might encounter, wishes for features, and so on. The answer for your problems may already lie in the forum archives, so searching the discussions is always the best way to begin.

Report Bugs

If you have found a possible bug in Vaadin, the demo applications, or the documentation, please report it by

Supplementary Material

xi


Preface

filing a ticket at the Vaadin developer's site at http://dev.vaadin.com/.You may want to check the existing tickets before filing a new one. You can make a ticket to make a request for a new feature as well, or to suggest modifications to an existing feature. Commercial Support

Vaadin offers full commercial support and training services for the Vaadin Framework and related products. Read more about the commercial products at http://vaadin.com/pro for details.

About the Author Marko Grรถnroos is a professional writer and software developer working at Vaadin Ltd in Turku, Finland. He has been involved in web application development since 1994 and has worked on several application development frameworks in C, C++, and Java. He has been active in many open source software projects and holds an M.Sc. degree in Computer Science from the University of Turku.

Acknowledgements Much of the book is the result of close work within the development team at Vaadin Ltd. Joonas Lehtinen, CEO of Vaadin Ltd, wrote the first outline of the book, which became the basis for the first two chapters. Since then, Marko Grรถnroos has become the primary author. The development team has contributed several passages, answered numerous technical questions, reviewed the manual, and made many corrections. The contributors are (in chronological order): Joonas Lehtinen Jani Laakso Marko Grรถnroos Jouni Koivuviita Matti Tahvonen Artur Signell Marc Englund Henri Sara

About Vaadin Ltd Vaadin Ltd is a Finnish software company specializing in the design and development of Rich Internet Applications. The company offers planning, implementation, and support services for the software projects of its customers, as well as sub-contract software development. Vaadin Framework, previously known as IT Mill Toolkit, is the flagship open source product of the company, for which it provides commercial development and support services.

xii

About the Author


Chapter 1

Introduction 1.1. Overview .................................................................................................... 1 1.2. Example Application Walkthrough ............................................................. 3 1.3. Support for the Eclipse IDE ....................................................................... 4 1.4. Goals and Philosophy ................................................................................ 4 1.5. Background ................................................................................................ 5 This chapter provides an introduction to software development with Vaadin, including installation of Vaadin, the Eclipse development environment, and any other necessary or useful utilities. We look into the design philosophy behind Vaadin, its history, and recent major changes.

1.1. Overview The core piece of Vaadin is the Java library that is designed to make creation and maintenance of high quality web-based user interfaces easy. The key idea in the server-driven programming model of Vaadin is that it allows you to forget the web and lets you program user interfaces much like you would program any Java desktop application with conventional toolkits such as AWT, Swing, or SWT. But easier. While traditional web programming is a fun way to spend your time learning new web technologies, you probably want to be productive and concentrate on the application logic. With the serverdriven programming model, Vaadin takes care of managing the user interface in the browser and AJAX communications between the browser and the server. With the Vaadin approach, you do not need to learn and debug browser technologies, such as HTML or JavaScript.

Book of Vaadin

1


Introduction

Figure 1.1. General Architecture of Vaadin

Figure 1.1, “General Architecture of Vaadin” illustrates the basic architecture of web applications made with Vaadin. Vaadin consists of the server-side framework and a client-side engine that runs in the browser as a JavaScript program, rendering the user interface and delivering user interaction to the server. As the application runs as a persistent Java Servlet session in an application server, you can easily bind your application to data and logic tiers. Because HTML, JavaScript, and other browser technologies are essentially invisible to the application logic, you can think of the web browser as only a thin client platform. A thin client displays the user interface and communicates user events to the server at a low level. The control logic of the user interface runs on a Java-based web server, together with your business logic. By contrast, a normal client-server architecture with a dedicated client application would include a lot of application specific communications between the client and the server. Essentially removing the user interface tier from the application architecture makes our approach a very effective one. As the Client-Side Engine is executed as JavaScript in the browser, no browser plugins are needed for using applications made with Vaadin. This gives it a sharp edge over frameworks based on Flash, Java Applets, or other plugins. Vaadin relies on the support of GWT for a wide range of browsers, so that the developer doesn't need to worry about browser support. Behind the server-driven development model, Vaadin makes the best use of AJAX (Asynchronous JavaScript and XML) techniques that make it possible to create Rich Internet Applications (RIA) that are as responsive and interactive as desktop applications. If you're a newcomer to AJAX, see Section 3.2.1, “AJAX” to find out what it is and how AJAX applications differ from traditional web applications. Hidden well under the hood, Vaadin uses GWT, the Google Web Toolkit, for rendering the user interface in the browser. GWT programs are written in Java, but compiled into JavaScript, thus freeing the developer from learning JavaScript and other browser technologies. GWT is ideal for implementing advanced user interface components (or widgets in GWT terminology) and interaction logic in the browser, while Vaadin handles the actual application logic in the server. Vaadin is designed to be extensible, and you can indeed use any 3rd-party GWT components easily, in addition to the component repertoire offered in Vaadin. The use of GWT also means that all the code you need to write is pure Java.

2

Overview


Introduction

The Vaadin library defines a clear separation between user interface presentation and logic and allows you to develop them separately. Our approach to this is themes, which dictate the visual appearance of applications. Themes control the appearance of the user interfaces using CSS and (optional) HTML page templates. As Vaadin provides excellent default themes, you do not usually need to make much customization, but you can if you need to. For more about themes, see Chapter 8, Themes. We hope that this is enough about the basic architecture and features of Vaadin for now. You can read more about it later in Chapter 3, Architecture, or jump straight to more practical things in Chapter 4, Writing a Web Application.

1.2. Example Application Walkthrough Let us follow the long tradition of first saying "Hello World!" when learning a new programming environment. After that, we can go through a more detailed example that implements the modelview-controller architecture. The two examples given are really simple, but this is mostly because Vaadin is designed to make things simple.

Example 1.1. HelloWorld.java import com.vaadin.ui.*; public class HelloWorld extends com.vaadin.Application { public void init() { Window main = new Window("Hello window"); setMainWindow(main); main.addComponent(new Label("Hello World!")); } }

The first thing to note is that the example application extends com.vaadin.Application class. The Application class is used as the base class for all user applications. Instances of the Application are essentially user sessions, and one is created for each user who uses the application. In the context of our HelloWorld application, it is sufficient to know that the application is started when the user first accesses it and at that time init method is invoked. Initialization of the application first creates a new window object and sets "Hello window" as its caption. The window is then set as the main window of the application; an application can actually have many windows. This means that when a user launches the application, the contents of the "main window" are shown to the user in the web page. The caption is shown as the title of the (browser) window. A new user interface component of class com.vaadin.ui.Label is created. The label is set to draw the text "Hello World!". Finally, the label is added to the main window. And here we are, when the application is started, it draws the text "Hello World!" to the browser window. The following screenshot shows what the "Hello World!" program will look like in a web browser.

Example Application Walkthrough

3


Introduction

Before going into details, we should note that this example source code is complete and does not need any additional declaratively defined template files to be run. To run the program, you can just add it to your web application, as explained in Section 4.8, “Setting Up the Application Environment”.

1.3. Support for the Eclipse IDE While Vaadin is not bound to any specific IDE, and you can in fact easily use it without any IDE altogether, we provide special support for the Eclipse IDE, which has become the standard environment for Java development. The support includes: • Import the installation package as a QuickStart demo project in Eclipse • Install the Vaadin Plugin for Eclipse, which allows you to: • Create new Vaadin projects • Create custom themes • Create custom client-side widgets and widget sets • Edit components with a visual (WYSIWYG) editor • Easily upgrade to a newer version of the Vaadin library The Vaadin Plugin for Eclipse is our recommended way of installing Vaadin; the actual installation package contains demos and documentation that are available also from the website, so you do not normally need to download and install it, unless you want to experiment with the demos or try debugging. Installing and updating the plugin is covered in Section 2.2.5, “Vaadin Plugin for Eclipse” and the creation of a new Vaadin project using the plugin in Section 2.4.1, “Creating the Project”. See Section 8.4, “Creating a Theme in Eclipse”, Section 10.2, “Doing It the Simple Way in Eclipse”, and Chapter 7, Visual User Interface Design with Eclipse (experimental) for instructions on using the different features of the plugin.

1.4. Goals and Philosophy Simply put, Vaadin's ambition is to be the best possible tool when it comes to creating web user interfaces for business applications. It is easy to adopt, as it is designed to support both entrylevel and advanced programmers, as well as usability experts and graphical designers. When designing Vaadin, we have followed the philosophy inscribed in the following rules.

4

Support for the Eclipse IDE


Introduction

Right tool for the right purpose Because our goals are high, the focus must be clear. This toolkit is designed for creating web applications. It is not designed for creating websites or advertisements demos. For such purposes, you might find (for instance) JSP/JSF or Flash more suitable.

Simplicity and maintainability We have chosen to emphasize robustness, simplicity, and maintainability. This involves following the well-established best practices in user interface frameworks and ensuring that our implementation represents an ideal solution for its purpose without clutter or bloat.

XML is not designed for programming The Web is inherently document-centered and very much bound to the declarative presentation of user interfaces. The Vaadin framework frees the programmer from these limitations. It is far more natural to create user interfaces by programming them than by defining them in declarative templates, which are not flexible enough for complex and dynamic user interaction.

Tools should not limit your work There should not be any limits on what you can do with the framework: if for some reason the user interface components do not support what you need to achieve, it must be easy to add new ones to your application. When you need to create new components, the role of the framework is critical: it makes it easy to create re-usable components that are easy to maintain.

1.5. Background The library was not written overnight. After working with web user interfaces since the beginning of the Web, a group of developers got together in 2000 to form IT Mill. The team had a desire to develop a new programming paradigm that would support the creation of real user interfaces for real applications using a real programming language. The library was originally called Millstone Library. The first version was used in a large production application that IT Mill designed and implemented for an international pharmaceutical company. IT Mill made the application already in the year 2001 and it is still in use. Since then, the company has produced dozens of large business applications with the library and it has proven its ability to solve hard problems easily. The next generation of the library, IT Mill Toolkit Release 4, was released in 2006. It introduced an entirely new AJAX-based presentation engine. This allowed the development of AJAX applications without the need to worry about communications between the client and the server.

Release 5 Into the Open IT Mill Toolkit 5, released initially at the end of 2007, took a significant step further into AJAX. The client-side rendering of the user interface was completely rewritten using GWT, the Google Web Toolkit. IT Mill Toolkit 5 introduced many significant improvements both in the server-side API and in the functionality. Rewriting the Client-Side Engine with GWT allowed the use of Java both on the client and the server-side. The transition from JavaScript to GWT made the development and integration of custom components and customization of existing components much easier than

Right tool for the right purpose

5


Introduction

before, and it also allows easy integration of existing GWT components. The adoption of GWT on the client-side did not, by itself, cause any changes in the server-side API, because GWT is a browser technology that is hidden well behind the API. Also themeing was completely revised in IT Mill Toolkit 5. The Release 5 was published under the Apache License 2, an unrestrictive open source license, to create faster expansion of the user base and make the formation of a developer community possible. Stabilization of the release 5 took over a year of work from the development team. It introduced a number of changes in the API, the client-side customization layer, and the themes. Many significant changes were done during the beta phase, until the stable version 5.3.0 was released in March 2009.

Birth of Vaadin Release 6 IT Mill Toolkit was renamed as Vaadin Framework, or Vaadin in short, in spring 2009. Later IT Mill, the company, was renamed as Vaadin Ltd. Vaadin means an adult female semi-domesticated mountain reindeer in Finnish. The most notable enhancements in Vaadin 6 are the external development tools: • Eclipse Plugin • Visual user interface editor under Eclipse (experimental) The Eclipse Plugin allows easy creation of Vaadin projects and custom client-side widgets. See Section 2.2.5, “Vaadin Plugin for Eclipse” for details. The visual editor, described in Chapter 7, Visual User Interface Design with Eclipse (experimental) makes prototyping easy and new users of Vaadin should find it especially useful for introducing oneself to Vaadin. Like Vaadin itself, the tools are open source. While the API in Vaadin 6 is essentially backward-compatible with IT Mill Toolkit 5.4, the package names and some name prefixes were changed to comply with the new product name: • Package name com.itmill.toolkit was renamed as com.vaadin. • The static resource directory ITMILL was changed to VAADIN. • Client-side widget prefix was changed from "I" to "V". • CSS style name prefix was changed from "i-" to "v-". Other major enhancements in Vaadin 6 are listed in the Release Notes for Vaadin 6.0.0, which also gives detailed instructions for upgrading from IT Mill Toolkit 5.

6

Birth of Vaadin Release 6


Chapter 2

Getting Started with Vaadin 2.1. Installing Vaadin ......................................................................................... 7 2.2. Setting up the Development Environment ............................................... 11 2.3. QuickStart with Eclipse ............................................................................ 18 2.4. Your First Project with Vaadin .................................................................. 24 This chapter gives practical instructions for installing Vaadin and the reference toolchain, installing the Vaadin plugin in Eclipse, and running and debugging the demo applications.

2.1. Installing Vaadin This section gives an overview of the Vaadin package and its installation. You have two options for installing: 1. Download and install the installation package 2. If you use Eclipse, you can install the Vaadin plugin for Eclipse, as described in Section 2.2.5, “Vaadin Plugin for Eclipse� Even if you use Eclipse, you can install the installation package and import it under Eclipse as the QuickStart project. It allows you to run and debug the demo applications and contains documentation.

Book of Vaadin

7


Getting Started with Vaadin

2.1.1. Installing the Distribution Package You can install the Vaadin installation package in a few simple steps: 1. Download the newest Vaadin installation package from the download page at http://vaadin.com/download/. 2. Unpack the installation package to a directory using a decompression program available in your operating system (see below). • In Windows, use the default ZIP decompression program to unpack the package into your chosen directory e.g. C:\dev. Warning At least the Windows XP default decompression program and some versions of WinRAR cannot unpack the installation package properly in certain cases. Decompression can result in an error such as "The system cannot find the file specified." This happens because the decompression program is unable to handle long file paths where the total length exceeds 256 characters. This occurs, for example, if you try to unpack the package under Desktop. You should unpack the package directly into C:\dev or some other short path or use another decompression program.

• In Linux, Mac OS X, and other UNIX-like systems, use Info-ZIP or other ZIP software with the unzip vaadin-6.x.x.zip command. The files will be, by default, decompressed into a directory with the name vaadin-6.x.x.

2.1.2. Starting the Content Browser The Content Browser is your best friend when using Vaadin. It allows you to browse documentation and example source code, and run the demo applications. The demo applications demonstrate most of the core features of Vaadin.You can find the demo applications also at the vaadin website: http://vaadin.com/demo. To start the Content Browser, run the start script in the Vaadin installation directory as instructed below. The start script launches a stand-alone web server running on the local host at port 8888, and a web browser at address http://localhost:8888/. The Content Browser will open the default web browser configured in your system. If your default browser is not compatible with Vaadin the demo applications may not work properly. In that case launch a supported browser manually and navigate to http://localhost:8888/. If the Content Browser fails to start, make sure that no other service is using port 8888. JRE must be installed You must have Java Runtime Environment (JRE) installed or the batch file will fail and close immediately. A JRE can be downloaded from http://java.sun.com/javase/downloads/index.jsp.

8

Installing the Distribution Package


Getting Started with Vaadin

Firewall software Executing the Content Browser locally may cause a security warning from your firewall software because of the started web server. You need to allow connections to port 8888 for the Content Browser to work.

Windows Run the start.bat batch file by double-clicking on the icon. Wait until the web server and web browser has started, it can take a while.

Linux / UNIX Open a shell window, change to the Vaadin installation directory, and run the start.sh shell script. You have to run it with the following command: $ sh start.sh ------------------------------------------------Starting Vaadin in Desktop Mode. Running in http://localhost:8888 ------------------------------------------------2007-12-04 12:44:55.657::INFO: Logging to STDERR via org.mortbay.log.StdErrLog 2007-12-04 12:44:55.745::INFO: jetty-6.1.5 2007-12-04 12:45:03.642::INFO: NO JSP Support for , did not find org.apache.jasper.servlet.JspServlet 2007-12-04 12:45:03.821::INFO: Started SelectChannelConnector@0.0.0.0:8888

Wait until the web server and web browser has started, it can take a while.

Mac OS X Double-click on the Start Vaadin icon. Wait until the web server and web browser has started, it can take a while. If the start icon fails in your environment for some reason, you can start the Content Browser by following the instructions for Linux/UNIX above: open a shell window, change to the installation directory, and execute sh start.sh. Browser support Vaadin supports the most commonly used web browsers, including Internet Explorer 6-8, Firefox 3, Safari 3 and Opera 9.6. In addition to these, most of the modern web browsers also work event if they are not supported. The definitive list of supported browsers can be found on http://vaadin.com/features.

2.1.3. Package Contents At the root of installation directory, you can find the start.bat (Windows) or start.sh (Linux and Mac) script that launches the Vaadin Content Browser, which allows you to run the demo applications and read the documentation included in the package.

Package Contents

9


Getting Started with Vaadin

Figure 2.1. Vaadin Package Contents

The WebContent directory contains all files related to Vaadin and the Content Browser. If you do not wish to or can not run the Content Browser, you can open index.html in a web browser to view the installation package contents and documentation. The demos will not be usable though. The release-notes.html file contains information about changes in the latest release and the release history. The license subdirectory contains licensing guidelines (licensing-guidelines.html) for Vaadin and licenses for other libraries included in the installation package. The COPYING file in the installation root also contains license information. The Vaadin Library itself is located at WebContent/vaadin-6.x.x.jar. The JAR package contains, in addition to the compiled files, the full source code of the library. The WebContent/docs directory contains full documentation for Vaadin, including JavaDoc API Reference Documentation (api subdirectory) and this manual in both HTML and printable PDF format. The WebContent/WEB-INF directory contains source code for the demo applications in the src subdirectory and the required libraries in the lib subdirectory. The gwt folder contains the full Google Web Toolkit installation package, including runtime libraries for the selected operating system, full documentation, and examples. You will need GWT if you intend to compile custom client-side widgets for Vaadin (described in Chapter 10, Developing Custom Components). In addition, the installation directory contains project files to allow importing the installation package as a project into the Eclipse IDE. See Section 2.3, “QuickStart with Eclipse� for details on how to do this.

10

Package Contents


Getting Started with Vaadin

2.1.4. Demo Applications The Content Browser allows you to run several demo applications included in the installation package. The applications demonstrate how you can use Vaadin for different tasks. Below is a selection of the included demos. Notice that the source code for all demos is included in the installation package and you can directly modify them if you import the package as a project in Eclipse, as instructed in Section 2.3, “QuickStart with Eclipse”. Sampler

Sampler demonstrates the various standard components and features of Vaadin. Clicking on one of the available sample icons will take you to the sample page, where you can see a live version of the sample together with a short description of feature. The Sampler allows you to view the full source code of each sample and provides links to the API documentation and to related samples. Sampler is the best place to get an overview of what is included in Vaadin.

Address Book Tutorial

This step-by-step tutorial covers everything needed to build a Vaadin application. The tutorial shows how you create layouts/views, implement navigation between views, bind components to a data source, use notifications, and much more. It also includes a section on how to create a custom theme for your application.

Reservation Application

The Reservation Application demonstrates the use of various components in a semi-real application connected to a local database. It also shows how to integrate a Google Maps view inside an application.

Coverflow

A simple example on how you can integrate Vaadin with Flex.

VaadinTunes

A non-functional application that demonstrates how you can create complex layouts using Vaadin.

Note: starting the demo applications can take several seconds.

2.2. Setting up the Development Environment This section gives a step-by-step guide for setting up a development environment.Vaadin supports a wide variety of tools, so you can use any IDE for writing the code, most web browsers for viewing the results, any operating system or processor supported by the Java 1.5 platform, and almost any Java server for deploying the results. In this example, we use the following toolchain: • Windows XP [http://www.microsoft.com/windowsxp/], Linux, or Mac OS X • Sun Java 2 Standard Edition 6.0 [http://java.sun.com/javase/downloads/index.jsp] (Java 1.5 or newer is required) • Eclipse IDE for Java EE Developers [http://www.eclipse.org/downloads/] • Apache Tomcat 6.0 (Core) or newer [http://tomcat.apache.org/]

Demo Applications

11


Getting Started with Vaadin

• Firefox 3.0.7 or newer [http://www.getfirefox.com/] • Firebug 1.3.3 or newer [http://www.getfirebug.com/] (optional) • Vaadin 6.x.x [http://vaadin.com/download/] The above is a good choice of tools, but you can use almost any tools you are comfortable with.

Figure 2.2. Development Toolchain and Process

Figure 2.2, “Development Toolchain and Process” illustrates the development environment and process.You develop your application as an Eclipse project. The project must include, in addition to your source code, the Vaadin Library. It can also include your project-specific themes. You must compile and deploy a project to a web container before you can use use it. You can deploy a project through the Web Tools Platform for Eclipse, which allows automatic deployment of web applications from Eclipse. You can deploy a project also manually, by creating a web application archive (WAR) and deploying it through the web container's interface.

2.2.1. Installing Java SDK Java SDK is required by Vaadin and also by the Eclipse IDE. Vaadin is compatible with Java 1.5 and later editions.

Windows 1. Download Sun Java 2 Standard Edition 6.0 from http://java.sun.com/javase/downloads/index.jsp [http://java.sun.com/javase/downloads/index.jsp]

12

Installing Java SDK


Getting Started with Vaadin

2. Install the Java SDK by running the installer. The default options are fine.

Linux / UNIX 1. Download Sun Java 2 Standard Edition 6.0 from http://java.sun.com/javase/downloads/index.jsp [http://java.sun.com/javase/downloads/index.jsp] 2. Decompress it under a suitable base directory, such as /opt. For example, for Java SDK, enter (either as root or with sudo in Linux): # cd /opt # sh (path-to-installation-package)/jdk-6u1-linux-i586.bin

and follow the instructions in the installer.

2.2.2. Installing Eclipse IDE Windows Eclipse is now installed in C:\dev\eclipse and can be started from there (by double clicking eclipse.exe). 1. Download Eclipse IDE for Java EE Developers (Ganymede version) from http://www.eclipse.org/downloads/ [http://www.eclipse.org/downloads/] 2. Decompress the Eclipse IDE package to a suitable directory. You are free to select any directory and to use any ZIP decompressor, but in this example we decompress the ZIP file by just double-clicking it and selecting "Extract all files" task from Windows compressed folder task. In our installation example, we use C:\dev as the target directory.

Linux / UNIX You have two basic options for installing Eclipse in Linux and UNIX: you can either install it using the package manager of your operating system or by downloading and installing the packages manually. The manual installation method is recommended, because the latest versions of the packages available in a Linux package repository may be incompatible with Eclipse plugins that are not installed using the package manager. 1. Download Download Eclipse IDE for Java EE Developers (Ganymede version) from http://www.eclipse.org/downloads/ [http://www.eclipse.org/downloads/] 2. Decompress the Eclipse package into a suitable base directory. It is important to make sure that there is no old Eclipse installation in the target directory. Installing a new version on top of an old one probably renders Eclipse unusable. 3. Eclipse should normally be installed as a regular user, as this makes installation of plugins easier. Eclipse also stores some user settings in the installation directory. To install the package, enter: $ tar zxf (path-to-installation-package)/eclipse-jee-ganymede-SR2-linux-gtk.tar.gz

This will extract the package to a subdirectory with the name eclipse. 4. You may wish to add the Eclipse installation directory and the bin subdirectory in the installation directory of Java SDK to your system or user PATH.

Installing Eclipse IDE

13


Getting Started with Vaadin

An alternative to the above procedure is to use the package management system of your operating system. For example, in Ubuntu Linux, which includes Sun Java SDK and Eclipse in its APT repository, you can install the programs from a package manager GUI or from command-line with a command such as: $ sudo apt-get install sun-java6-jdk eclipse

This is, however, not recommended, because the Eclipse package may not include all the necessary Java EE tools, most importantly the Web Standard Tools, and it may cause incompatibilities with some components that are not installed with the package management system of your operating system.

2.2.3. Installing Apache Tomcat Apache Tomcat is a lightweight Java web server suitable for both development and production. There are many ways to install it, but here we simply decompress the installation package. Apache Tomcat should be installed with user permissions. During development, you will be running Eclipse or some other IDE with user permissions, but deploying web applications to a Tomcat server that is installed system-wide requires administrator or root permissions. 1. Download the installation package: Apache Tomcat 6.0 (Core Binary Distribution) from http://tomcat.apache.org/ 2. Decompress Apache Tomcat package to a suitable target directory, such as C:\dev (Windows) or /opt (Linux or Mac OS X). The Apache Tomcat home directory will be C:\dev\apache-tomcat-6.0.x or /opt/apache-tomcat-6.0.x, respectively.

2.2.4. Firefox and Firebug Vaadin supports many web browsers and you can use any of them for development. If you plan to create a custom theme, customized layouts or new user interface components, we recommend that you use Firefox together with Firebug for debugging. Vaadin contains special support for Firebug and can show debug information in its console. If you do not have Firefox installed already, go to www.getfirefox.com [http://www.getfirefox.com/] and download and run the installer. Optional. After installing Firefox, use it to open http://www.getfirebug.com/ [http://www.getfirebug.com/]. Follow the instructions on the site to install the latest stable version of Firebug available for the browser. You might need to tell Firefox to allow the installation by clicking the yellow warning bar at the top of the browser-window. When Firebug is installed, it can be enabled at any time from the bottom right corner of the Firefox window. Figure 2.3, “Firebug Debugger for Firefox� shows an example of what Firebug looks like.

14

Installing Apache Tomcat


Getting Started with Vaadin

Figure 2.3. Firebug Debugger for Firefox

Now that you have installed the development environment, you can proceed to creating your first application.

2.2.5. Vaadin Plugin for Eclipse If you are using the Eclipse IDE, using the Vaadin plugin should help greatly. The plugin includes: • An integration plugin with wizards for creating new Vaadin-based projects, themes, and client-side widgets and widget sets. • A visual editor for editing custom composite user interface components in a WYSIWYG fashion. With full round-trip support from source code to visual model and back, the editor integrates seamlessly with your development process. • A version of Book of Vaadin that you can browse in the Eclipse Help system. You can install the plugin as follows: 1. Start Eclipse. 2. Select Help → Software Updates.... 3. Select the Available Software tab. 4. Add the Vaadin plugin update site by clicking Add Site....

Vaadin Plugin for Eclipse

15


Getting Started with Vaadin

Enter the URL of the Vaadin Update Site: http://vaadin.com/eclipse and click OK. The Vaadin site should now appear in the Software Updates window. 5. Select all the Vaadin plugins in the tree.

Finally, click Install. Detailed and up-to-date installation instructions for the Eclipse plugin can be found at http://vaadin.com/eclipse.

Updating the Vaadin Plugin If you have automatic updates enabled in Eclipse (see Window → Preferences → Install/Update → Automatic Updates), the Vaadin plugin will be updated automatically along with other plugins. Otherwise, you can update the Vaadin plugin (there are actually multiple plugins) manually as follows: 1. Select Help → Software Updates..., the Software Updates and Add-ons window will open. 2. Select the Installed Software tab. 3. If you want to update only the Vaadin plugins, select them in the list by clicking the plugins and holding the Ctrl key pressed for all but the first.

16

Vaadin Plugin for Eclipse


Getting Started with Vaadin

4. Click Update. Notice that updating the Vaadin plugin updates only the plugin and not the Vaadin library, which is project specific. See below for instructions for updating the library.

Updating the Vaadin Library Updating the Vaadin plugin does not update Vaadin library. The library is project specific, as a different version might be required for different projects, so you have to update it separately for each project. To change the library to a newer (or some other) version, do as follows: 1. Select the project in the Project Explorer and select Project → Preferences or press Alt-Enter. 2. In the project preferences window that opens, select Vaadin → Vaadin Version. 3. If the version that you want to use is not included in the Vaadin version drop-down list, click Download to open the download window.

If you want to use a development version, select Show pre-release versions and nightly builds. Select the version that you want to download and click OK.

Vaadin Plugin for Eclipse

17


Getting Started with Vaadin

4. Select the version that you want to use from the Vaadin version down-down list and click Apply. You can observe that the new library appears in the WebContent/WEB-INF/lib folder.

2.3. QuickStart with Eclipse Eager to start developing you own applications using Vaadin? This section presents a QuickStart into running and debugging Vaadin demos under Eclipse. The QuickStart includes a web server, so you do not need to have a full-weight web container such as Apache Tomcat installed.

2.3.1. Starting Eclipse If you have followed the instructions given in Section 2.2, “Setting up the Development Environment”, you can start Eclipse by running C:\dev\eclipse\eclipse.exe (Windows) or /opt/eclipse/eclipse (Linux or OS X). Depending on your environment, you may need to give additional memory settings for Eclipse, as the default values are known to cause problems often in some systems. When starting Eclipse for the first time, it asks where to save the workspace. You can select any directory, but here we select C:\dev\workspace (Windows) or /home/<user>/workspace (Linux or OS X). We suggest that you also set this as the default. Close the Eclipse "Welcome" -screen when you are ready to continue.

2.3.2. Importing Vaadin as a Project If you have not yet downloaded the Vaadin package, instructions for downloading and unpacking are available at Section 2.1.1, “Installing the Distribution Package”. The installation directory of Vaadin contains all necessary files to allow importing it as a readyto-run Eclipse project: 1. Start Eclipse with any workspace you like. Switch to the Java Perspective through Window → Open Perspective → Java. 2. Select File → Import... to open the import dialog. 3. In the Import dialog, select General → Existing Projects into Workspace and click Next. 4. In the Select root directory option, click the Browse button, and select the folder where you unpacked Vaadin, such as, c:/dev/vaadin-6.x.x. Click OK in the selection window. The Projects list now shows a project named vaadin-examples. Click Finish in the Import window to finish importing the project. The newly imported project should look like Figure 2.4, “Vaadin demo project imported into Eclipse”.

18

QuickStart with Eclipse


Getting Started with Vaadin

Figure 2.4. Vaadin demo project imported into Eclipse

You can now browse the source code of the demo applications in Eclipse. The next section describes how you can run the demos.

2.3.3. Running the Demo Applications in Eclipse Once the project is imported, as described above, you can run the Content Browser, including the demo applications, as follows: 1. From the main menu, select Run â&#x2020;&#x2019; Run Configurations....

Running the Demo Applications in Eclipse

19


Getting Started with Vaadin 2. From the list on the left, select Java Application â&#x2020;&#x2019; Vaadin Web Mode. 3. Click Run. Note that after the application has been launched once, it appears in the Favorites list. You can then click on the small dropdown arrow on the right side of the Run button on Eclipse toolbar and select Vaadin Development Server.

Running the application in Vaadin Development Server will start an embedded web server and open a browser window with the Content Browser. The default system web browser is opened; make sure that the browser is compatible with Vaadin. The Console view in the lower pane of Eclipse will display text printed to standard output by the application. Clicking on the red Terminate button will stop the server.

20

Running the Demo Applications in Eclipse


Getting Started with Vaadin

Figure 2.5. Vaadin Content Browser Started Under Eclipse

Note that executing the web application locally may cause a security warning from your firewall software because of the started web server. You need to allow connections to port 8888 for the Content Browser to work. Also, if the web service fails to start, make sure that no other service is using port 8888.

Launching the GWT Development Mode The Google Web Toolkit Development Mode allows debugging client-side GWT Java code in an IDE such as Eclipse. This is made possible by the Google Web Toolkit Developer Plugin, which

Running the Demo Applications in Eclipse

21


Getting Started with Vaadin

is available for major browsers, such as Mozilla Firefox, Google Chrome, Safari, and Internet Explorer. To run the demo applications in the GWT Development Mode, follow the following steps: 1. If not already started, start the demo application in Vaadin Development Server as described above. We only need the server so close the web browser which is automatically opened. 2. From the main menu, select Run → Debug Configurations... . 3. From the list select Java Application → GWT Development Mode. 4. Click Debug. This will open the GWT Development Mode window. 5. Click Launch Default Browser. This will open the application page in the default browser with a gwt.codesvr parameter, which will make the application use the GWT Development Mode. If you have not yet installed the GWT Developer Plugin, the browser will open an installation page, which allows you to download and install the plugin. Please follow the instructions. After the installation is complete (the browser may need to be restarted), the application will be started. Starting demo applications under the GWT Development Mode can take considerable time! Starting the development mode compiles the widget set of the application, which can take 10-60 seconds, depending on the hardware. During this time, the web browser is completely unresponsive and does not update its window - you should not think that the browser is stuck. As with the Vaadin Development Server launcher, after you have run the GWT Development Mode launcher once, you can click the dropdown marker on right of the Debug button in the toolbar and find the GWT Development Mode in the dropdown list. To use the GWT Development Mode in other projects, you need to create a launch configuration in Eclipse. See Section 10.8.6, “GWT Development Mode” for more detailed information about the Development Mode and how to create the launch configuration.

How to Stop the Run To stop the launched Jetty web container that serves the Content Browser web application, select the Console tab and click on the Terminate button.

Figure 2.6. Terminating a Launch

To clean up all terminated launches from the Console window, click on the Remove All Terminated Launches button.

Figure 2.7. Removing Terminated Launches

22

Running the Demo Applications in Eclipse


Getting Started with Vaadin

2.3.4. Debugging the Demo Applications in Eclipse At some point when developing an application, you want to debug it. Running a web application in debug mode is easy in Eclipse. Next, we will show you how to debug the demo applications by inserting a breakpoint in the Calc example. 1. Make sure to stop any previous Run command as instructed above at the end of Section 2.3.3, “Running the Demo Applications in Eclipse”. 2. Select Run → Debug Configurations... from the main menu and the Debug configuration window will open. 3. Select Java Application → Vaadin Web Mode and click Debug. The server will start and the web browser will open. 4. Open the source code for the Calc program. It is located in WebContent/WEB-INF/src/com.vaadin.demo.Calc. WebContent/WEB-INF/src is the project's source folder, shown right below the JRE System Library. Double-click the class to open the source code in the editor. 5. Insert a breakpoint in the init() by right-clicking on the gray bar on the left of the editor window to open the context menu, and select Toggle Breakpoint. 6. Switch to the browser window and click the Calc link (below More Examples) to open it. 7. Eclipse encouters the breakpoint and asks to switch to the Debug perspective. Click Yes. The debug window will show the current line where the execution stopped as illustrated in Figure 2.8, “Execution Stopped at Breakpoint in Debug Perspective in Eclipse”:

Figure 2.8. Execution Stopped at Breakpoint in Debug Perspective in Eclipse

8. You can now step forward or use any commands you would normally use when debugging an application in Eclipse. Note that you can only debug the application code in this way. If you are debugging the application with the GWT Development Mode, you can also insert break-points in client side component code and debug it.

2.3.5. Using QuickStart as a Project Skeleton If you like, you can also use the imported Vaadin demo project as a skeleton for your own project. Just remove any unnecessary files or files related to the demo applications from the project. The proper way of creating a new Vaadin project will be described in the next section: Section 2.4, “Your First Project with Vaadin”.

Debugging the Demo Applications in Eclipse

23


Getting Started with Vaadin

2.4. Your First Project with Vaadin This section gives instructions for creating a new Eclipse project using the Vaadin Plugin. The task will include the following steps: 1. Create a new project 2. Write the source code 3. Configure and start Tomcat (or some other web server) 4. Open a web browser to use the web application We also show how you can debug the application in the debug mode in Eclipse. This walkthrough assumes that you have already installed the Vaadin Plugin and set up your development environment, as instructed in Section 2.1.1, “Installing the Distribution Package” and Section 2.2, “Setting up the Development Environment”.

2.4.1. Creating the Project Let us create the first application project with the tools installed in the previous section. First, launch Eclipse and follow the following steps: 1. Start creating a new project by selecting from the menu File → New → Project.... 2. In the New Project window that opens, select Web → Vaadin Project and click Next.

24

Your First Project with Vaadin


Getting Started with Vaadin

3. In the Vaadin Project step, you need to set the basic web project settings. You need to give at least the project name and the runtime; the default values should be good for the other settings.

Project name

Give the project a name. The name should be a valid identifier usable cross-platform as a filename and inside a URL, so using only lower-case alphanumerics, underscore, and minus sign is recommended.

Use default

Defines the directory under which the project is created. You should normally leave it as it is. You may need to set the directory, for example, if you are creating an Eclipse project on top of a version-controlled source tree.

Target runtime

Defines the application server to use for deploying the application. The server that you have installed, for example Apache Tomcat, should be selected automatically. If not, click New to configure a new server under Eclipse.

Configuration

Select the configuration to use; you should normally use the default configuration for the application server. If you need to modify the project facets, click Modify.

Deployment configuration

This setting defines the environment to which the application will be deployed, to generate the appropriate project directory layout and configuration files. The choises are: â&#x20AC;˘ Servlet (default),

Creating the Project

25


Getting Started with Vaadin

• Google App Engine Servlet, and • Generic Portlet (Portlet 2.0). • Old Portlet (Portlet 1.0). The further steps in the New Project Wizard depend on the selected deployment configuration; the steps listed in this section are for the default servlet configuration. See Section 11.9, “Google App Engine Integration” and Section 11.8, “Portal Integration” for details on the use of Vaadin in the alternative environments. Vaadin version

Select the Vaadin version to use. The drop-down list shows, by default, the latest available version of Vaadin. If you want to use another version, click Download. The dialog that opens lists all official releases of Vaadin.

If you want to use a pre-release version or a nightly development build, select Show pre-release versions and nightly builds. Select a version and click Ok to download it. It will appear as a choise in the drop-down list. If you want to change the project to use another version of Vaadin, for example to upgrade to a newer one, you can go to project settings and download and select the other version. You can click Finish here to use the defaults for the rest of the settings, or click Next.

26

Creating the Project


Getting Started with Vaadin

4. The settings in the Web Module step define the basic servlet-related settings and the structure of the web application project. All the settings are pre-filled, and you should normally accept them as they are.

Context Root

The context root (of the application) identifies the application in the URL used for accessing it. For example, if the server runs in the apps context and the application has myproject context, the URL would be http://example.com/app/url. The wizard will suggest myproject for the context name.

Content Directory

The directory containing all the content to be included in the servlet and served by the web server. The directory is relative to the root directory of the project.

Java Source Directory

The default source directory containing the application sources. The src directory is suggested; another convention common in web applications is to use WebContent/WEB-INF/src, in which case the sources are included in the servlet (but not served in HTTP requests).

Generate deployment descriptor

Should the wizard generate the web.xml deployment descriptor required for running the servlet in the WebContent/WEB-INF directory. Strongly recommended. See Section 4.8.3, â&#x20AC;&#x153;Deployment Descriptor web.xmlâ&#x20AC;? for more details.

This will be the sub-path in the URL, for example http://localhost:8080/myproject. The default for the application root will be / (root). You can just accept the defaults and click Next.

Creating the Project

27


Getting Started with Vaadin

5. The Vaadin project step page has various Vaadin-specific application settings. If you are trying Vaadin out for the first time, you should not need to change anything. You can set most of the settings afterwards, except the creation of the portlet configuration.

Create project template

Make the wizard create an application class stub.

Application Name

The name of the application appears in the browser window title.

Base package name

The name of the Java package under which the application class is to be placed.

Application class name

Name of the Vaadin application class.

Create portlet configuration

When this option is selected, the wizard will create the files needed for running the application in a portal. See Section 11.8, â&#x20AC;&#x153;Portal Integrationâ&#x20AC;? for more information on portlets.

Finally, click Finish to create the project. 6. Eclipse may ask to switch to J2EE perspective. A Dynamic Web Project uses an external web server and the J2EE perspective provides tools to control the server and manage application deployment. Click Yes.

2.4.2. Exploring the Project After the New Project wizard exists, it has done all the work for us: Vaadin libraries are installed in the WebContent/WEB-INF/lib directory, an application class skeleton has been written to src directory, and WebContent/WEB-INF/web.xml already contains a deployment descriptor.

28

Exploring the Project


Getting Started with Vaadin

Figure 2.9. A New Dynamic Web Project

The application class created by the plugin contains the following code: package com.example.myproject; import com.vaadin.Application; import com.vaadin.ui.*; public class MyprojectApplication extends Application { @Override public void init() { Window mainWindow = new Window("Myproject Application"); Label label = new Label("Hello Vaadin user"); mainWindow.addComponent(label); setMainWindow(mainWindow); } }

Let us add a button to the application to make it a bit more interesting. The resulting init() method could look something like: public void init() { final Window mainWindow = new Window("Myproject Application"); Label label = new Label("Hello Vaadin user"); mainWindow.addComponent(label); mainWindow.addComponent( new Button("What is the time?", new Button.ClickListener() { public void buttonClick(ClickEvent event) { mainWindow.showNotification( "The time is " + new Date()); } }));

Exploring the Project

29


Getting Started with Vaadin

setMainWindow(mainWindow); }

The deployment descriptor WebContent/WEB-INF/web.xml defines Vaadin framework servlet, the application class, and servlet mapping:

Example 2.1. Web.xml Deployment Descriptor for our project <?xml version="1.0" encoding="UTF-8"?> <web-app xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://java.sun.com/xml/ns/javaee" xmlns:web="http://java.sun.com/xml/ns/javaee/web-app_2_5.xsd" id="WebApp_ID" version="2.5"> <display-name>myproject</display-name> <context-param> <description>Vaadin production mode</description> <param-name>productionMode</param-name> <param-value>false</param-value> </context-param> <servlet> <servlet-name>Myproject Application</servlet-name> <servlet-class> com.vaadin.terminal.gwt.server.ApplicationServlet </servlet-class> <init-param> <description>Vaadin application class to start</description> <param-name>application</param-name> <param-value> com.example.myproject.MyprojectApplication </param-value> </init-param> </servlet> <servlet-mapping> <servlet-name>Myproject Application</servlet-name> <url-pattern>/*</url-pattern> </servlet-mapping> </web-app>

For a more detailed treatment of the web.xml file, see Section 4.8.3, “Deployment Descriptor web.xml”.

2.4.3. Setting Up and Starting the Web Server Eclipse IDE for Java EE Developers has the Web Standard Tools package installed, which supports control of various web servers and automatic deployment of web content to the server when changes are made to a project. Make sure that Tomcat was installed with user permissions. Configuration of the web server in Eclipse will fail if the user does not have write permissions to the configuration and deployment directories under the Tomcat installation directory. Follow the following steps. 1. Switch to the Servers tab in the lower panel in Eclipse. List of servers should be empty after Eclipse is installed. Right-click on the empty area in the panel and select New → Server.

30

Setting Up and Starting the Web Server


Getting Started with Vaadin

2. Select Apache â&#x2020;&#x2019; Tomcat v6.0 Server and set Server's host name as localhost, which should be the default. If you have only one Tomcat installed, Server runtime has only one choice. Click Next.

3. Add your project to the server by selecting it on the left and clicking Add to add it to the configured projects on the right. Click Finish.

4. The server and the project are now installed in Eclipse and are shown in the Servers tab. To start the server, right-click on the server and select Debug. To start the server in non-debug mode, select Start.

5. The server starts and the WebContent directory of the project is published to the server on http://localhost:8080/myproject/.

Setting Up and Starting the Web Server

31


Getting Started with Vaadin

2.4.4. Running and Debugging Starting your application is as easy as selecting myproject from the Project Explorer and then Run â&#x2020;&#x2019; Debug As â&#x2020;&#x2019; Debug on Server. Eclipse then opens the application in built-in web browser.

Figure 2.10. Running a Vaadin Application

You can insert break points in the Java code by double-clicking on the left margin bar of the source code window. For example, if you insert a breakpoint in the buttonClick() method and click the What is the time? button, Eclipse will ask to switch to the Debug perspective. Debug perspective will show where the execution stopped at the breakpoint. You can examine and change the state of the application. To continue execution, select Resume from Run menu.

Figure 2.11. Debugging a Vaadin Application

32

Running and Debugging


Getting Started with Vaadin

The procedure described above allows debugging the server-side application. For more information on debugging client-side widgets, see Section 10.8.6, â&#x20AC;&#x153;GWT Development Modeâ&#x20AC;?.

Running and Debugging

33


34


Chapter 3

Architecture 3.1. Overview .................................................................................................. 35 3.2. Technological Background ....................................................................... 38 3.3. Applications as Java Servlet Sessions .................................................... 39 3.4. Client-Side Engine ................................................................................... 39 3.5. Events and Listeners ............................................................................... 41 This chapter provides an introduction to the architecture of Vaadin at somewhat technical level.

3.1. Overview In Chapter 1, Introduction, we gave a short introduction to the general architecture of Vaadin. Let us now look deeper into it. Figure 3.1, â&#x20AC;&#x153;Vaadin Architectureâ&#x20AC;? below illustrates the architecture. Vaadin consists of a web application API, a horde of user interface components, themes for controlling the appearance, and a data model that allows binding the user interface components directly to data. Behind the curtains it also employs a terminal adapter to receive requests from web browsers and make responses by rendering the pages. An application using Vaadin runs as a servlet in a Java web server, serving HTTP requests. The terminal adapter receives client requests through the web server's Java Servlet API, and inteprets them to user events for a particular session. An event is associated with a UI component and delivered to the application. As the application logic makes changes to the UI components, the terminal adapter renders them in the web browser by generating a response. In AJAX rendering mode, a client-side JavaScript component receives the responses and uses them to make any necessary changes to the page in the browser. The top level of a user application consists of an application class that inherits com.vaadin.Application. It creates the UI components (see below) it needs, receives events regarding them,

Book of Vaadin

35


Architecture

Figure 3.1. Vaadin Architecture

and makes necessary changes to the components. For detailed information about inheriting the Application, see Chapter 4, Writing a Web Application. The major parts of the architecture and their function are as follows:

36

User Interface Components

The user interface consists of UI components that are created and laid out by the application. Each serverside component has a client-side counterpart, with which the user interacts. The server-side components can serialize themselves over the client connection using a terminal adapter. The client-side components, in turn, can serialize user interaction back to the application, which is received in the server-side components as events. The components relay these events to the application logic. Most components are bound to a data source (see below). For a complete description of UI component architecture, see Chapter 5, User Interface Components.

Client-Side Engine

The Client-Side Engine of Vaadin manages the rendering in the web browser using Google Web Toolkit (GWT). It communicates user interaction and UI changes with the server-side Terminal Adapter using

Overview


Architecture

the User Interface Definition Language (UIDL), a JSONbased language. The communications are made using asynchronous HTTP or HTTPS requests. See Section 3.4, “Client-Side Engine”. Terminal Adapter

The UI components do not render themselves directly as a web page, but use a Terminal Adapter. This abstraction layer allows users to use Vaadin applications with practically any web browser. Releases 3 and 4 of IT Mill Toolkit supported HTML and simple AJAX based rendering, while Vaadin Release 5 supports advanced AJAX-based rendering using Google Web Toolkit (GWT). You could imagine some other browser technology, not even based on HTML, and you - or we for that matter - could make it work just by writing a new adapter.Your application would still just see the Vaadin API. To allow for this sort of abstraction, UI components communicate their changes to the Terminal Adapter, which renders them for the user's browser. When the user does something in the web page, the events are communicated to the terminal adapter (through the web server) as asynchronous AJAX requests. The terminal adapter delivers the user events to the UI components, which deliver them to the application's UI logic.

Themes

The user interface separates between presentation and logic. While the UI logic is handled as Java code, the presentation is defined in themes as CSS. Vaadin provides a default themes. User themes can, in addition to style sheets, include HTML templates that define custom layouts and other resources, such as images. Themes are discussed in detail in Chapter 8, Themes.

UIDL

The Terminal Adapter draws the user interface to the web page and any changes to it using a special User Interface Definition Language (UIDL). The UIDL communications are done using JSON (JavaScript Object Notation), which is a lightweight data interchange format that is especially efficient for interfacing with JavaScriptbased AJAX code in the browser. See Section 3.2.3, “JSON” and Appendix A, User Interface Definition Language (UIDL) for details.

Events

User interaction with UI components creates events, which are first processed on the client-side with JavaScript and then passed all the way through the HTTP server, terminal adapter, and user component layers to the application. See Section 3.5, “Events and Listeners”.

Data Model

In addition to the user interface model, Vaadin provides a data model for interfacing data presented in UI components. Using the data model, the user interface components can update the application data directly,

Overview

37


Architecture

without the need for any control code. All the UI components use this data model internally, but they can be bound to a separate data source as well. For example, you can bind a table component to an SQL query response. For a complete overview of the Vaadin Data Model, please refer to Chapter 9, Binding Components to Data.

3.2. Technological Background This section provides an introduction to the various technologies and designs on which Vaadin is based: AJAX-based web applications in general, Google Web Toolkit, and JSON data interchange format. This knowledge is not necessary for using Vaadin, but provides some background if you need to make low-level extensions to Vaadin.

3.2.1. AJAX AJAX (Asynchronous JavaScript and XML) is a technique for developing web applications with responsive user interaction, similar to traditional desktop applications. While conventional JavaScript-enabled HTML pages can receive new content only with page updates, AJAX-enabled pages send user interaction to the server using an asynchronous request and receive updated content in the response. This way, only small parts of the page data can be loaded. This goal is archieved by the use of a certain set of technologies: XHTML, CSS, DOM, JavaScript, XMLHttpRequest, and XML. AJAX, with all the fuss and pomp it receives, is essentially made possible by a simple API, namely the XMLHttpRequest class in JavaScript. The API is available in all major browsers and, as of 2006, the API is under way to become a W3C standard. Communications between the browser and the server usually require some sort of serialization (or marshalling) of data objects. AJAX suggests the use of XML for data representation in communications between the browser and the server. While Vaadin Release 4 used XML for data interchange, Release 5 uses the more efficient JSON. For more information about JSON and its use in Vaadin, see Section 3.2.3, “JSON”. If you're a newcomer to Ajax, Section 11.1, “Special Characteristics of AJAX Applications” discusses the history and motivations for AJAX-based web applications, as well as some special characteristics that differ from both traditional web applications and desktop applications.

3.2.2. Google Web Toolkit Google Web Toolkit is a software development kit for developing client-side web applications easily, without having to use JavaScript or other browser technologies directly. Applications using GWT are developed with Java and compiled into JavaScript with the GWT Compiler. GWT is essentially a client-side technology, normally used to develop user interface logic in the web browser. GWT applications still need to communicate with a server using RPC calls and by serializing any data. Vaadin effectively hides all client-server communications, allows handling user interaction logic in a server application, and allows software development in a single serverside application. This makes the architecture of an AJAX-based web application much simpler. Vaadin uses GWT to render user interfaces in the web browser and handle the low-level tasks of user interaction in the browser. Use of GWT is largely invisible in Vaadin for applications that do not need any custom GWT components.

38

Technological Background


Architecture

See Section 3.4, “Client-Side Engine” for a description of how GWT is used in the Client-Side Engine of Vaadin. Chapter 10, Developing Custom Components provides information about the integration of GWT-based user interface components with Vaadin.

3.2.3. JSON JSON is a lightweight data-interchange format that is easy and fast to generate and parse. JSON messages are said to be possibly a hundred times faster to parse than XML with current browser technology. The format is a subset of the JavaScript language, which makes it possible to evaluate JSON messages directly as JavaScript expressions. This makes JSON very easy to use in JavaScript applications and therefore also for AJAX applications. The Client-Side Engine of Vaadin uses JSON through Google Web Toolkit, which supports JSON communications in the com.google.gwt.json.client package. Together with advanced update optimization and caching, Vaadin is able to update changes in the user interface to the browser in an extremely efficient way. The use of JSON is completely invisible to a developer using Vaadin. Implementation of clientserver serialization in custom widgets uses abstract interfaces that may be implemented as any low-level interchange format, such as XML or JSON. Details on JSON communications are given in Section A.2, “JSON Rendering”.

3.3. Applications as Java Servlet Sessions Vaadin framework does basically everything it does on top of the Java Servlet API, which lies hidden deep under the hood, with the terminal adapter being the lowest level layer for handling requests from the web container. When the web container gets the first request for a URL registered for an application, it creates an instance of the ApplicationServlet class in Vaadin framework that inherits the HttpServlet class defined in Java Servlet API. It follows sessions by using HttpSession interface and associates an Application instance with each session. During the lifetime of a session, the framework relays user actions to the proper application instance, and further to a user interface component.

3.4. Client-Side Engine This section gives an overview of the client-side architecture of Vaadin. Knowledge of the clientside technologies is generally not needed unless you develop or use custom GWT components. The client-side engine is based on Google Web Toolkit (GWT), which allows the development of the engine and client-side components solely with Java. Chapter 10, Developing Custom Components provides information about the integration of GWTbased user interface components with Vaadin.

JSON

39


Architecture

Figure 3.2. Architecture of Vaadin Client-Side Engine

Figure 3.2, â&#x20AC;&#x153;Architecture of Vaadin Client-Side Engineâ&#x20AC;? illustrates the architecture of the clientside engine using a button component as an example. The user interface is managed by the ApplicationConnection class, which handles AJAX requests to the server and renders the user interface according to responses. Communications are done over HTTP(S) using the JSON data interchange format and the User Interface Definition Language (UIDL). In the server-side application, the button is used with the Button class of Vaadin. On the client-side, the user interface consists of various GWT components that inherit Widget class. In the figure above, the GWT class Button is used to render the button in the browser (the inheritance of Button is simplified in the figure). Vaadin provides an VButton class, which implements the Paintable interface needed for rendering the component with GWT.

40

Client-Side Engine


Architecture

The actual initial web page that is loaded in the browser is an empty page that loads the JavaScript code of the Vaadin Client-Side Engine. After it is loaded and started, it handles the AJAX requests to the server. All server communications are done through the ApplicationConnection class. The communication with the server is done as UIDL (User Interface Definition Language) messages using the JSON message interchange format over a HTTP(S) connection. UIDL is described in Appendix A, User Interface Definition Language (UIDL) and JSON in Section 3.2.3, “JSON” and Section A.2, “JSON Rendering”.

3.5. Events and Listeners When a user does something, such as clicks a button or selects an item, the application needs to know about it. Many Java-based user interface frameworks follow the Observer design pattern to communicate user input to the application logic. So does Vaadin. The design pattern involves two kinds of elements: an object and a number of observers that listen for events regarding the object. When an event related to the object occurs, the observers receive a notification regarding the event. In most cases there is only one observer, defined in the application logic, but the pattern allows for multiple observers. As in the event-listener framework of Java SE, we call the observing objects listeners. In the ancient times of C programming, callback functions filled largely the same need as listeners do now. In object-oriented languages, we have only classes and methods, not functions, so the application has to give a class interface instead of a callback function pointer to the framework. However, Vaadin supports defining a method as a listener as well. Events can serve many kinds of purposes. In Vaadin, the usual purpose of events is handling user interaction in a user interface. Session management can require special events, such as time-out, in which case the event is actually the lack of user interaction. Time-out is a special case of timed or scheduled events, where an event occurs at a specific date and time or when a set time has passed. Database and other asynchronous communications can cause events too. To receive events of a particular type, an application must include a class that implements the corresponding listener interface. In small applications, the application class itself could implement the needed listener interfaces. Listeners are managed by the AbstractComponent class, the base class of all user interface components. This means that events regarding any component can listened to. The listeners are registered in the components with addListener() method. Most components that have related events define their own event class and corresponding listener classes. For example, the Button has Button.ClickEvent events, which can be listened to through the Button.ClickListener interface.This allows an application to listen to many different kinds of events and to distinguish between them at class level. This is usually not enough, as applications usually have many components of the same class and need to distinguish between the particular components. We will look into that more closely later. The purpose of this sort of class level separation is to avoid having to make type conversions in the handlers. Notice that many listener interfaces inherit the java.util.EventListener superinterface, but it is not generally necessary to inherit it.

Events and Listeners

41


Architecture

Figure 3.3. Class Diagram of a Button Click Listener

Figure 3.3, “Class Diagram of a Button Click Listener” illustrates an example where an applicationspecific class inherits the Button.ClickListener interface to be able to listen for button click events. The application must instantiate the listener class and register it with addListener(). When an event occurs, an event object is instantiated, in this case a ClickEvent.The event object knows the related UI component, in this case the Button. Section 4.4, “Handling Events with Listeners” goes into details of handling events in practice.

42

Events and Listeners


Chapter 4

Writing a Web Application 4.1. Overview .................................................................................................. 43 4.2. Managing the Main Window .................................................................... 46 4.3. Child Windows ......................................................................................... 46 4.4. Handling Events with Listeners ................................................................ 50 4.5. Referencing Resources ........................................................................... 52 4.6. Shutting Down an Application .................................................................. 56 4.7. Handling Errors ........................................................................................ 57 4.8. Setting Up the Application Environment .................................................. 61 This chapter provides the fundamentals of web application development with Vaadin, concentrating on the basic elements of an application from a practical point-of-view. If you are a newcomer to AJAX development, you may benefit from Section 11.1, â&#x20AC;&#x153;Special Characteristics of AJAX Applicationsâ&#x20AC;?. It explains the role of pages in AJAX web applications, and provides some basic design patterns for applications.

4.1. Overview An application made with Vaadin runs as a Java Servlet in a Servlet container. The entry-point is the application class, which needs to create and manage all necessary user interface components, including windows. User interaction is handled with event listeners, simplified by binding user interface components directly to data. Visual appearance is defined in themes as CSS files.

Book of Vaadin

43


Writing a Web Application

Icons, other images, and downloadable files are handled as resources, which can be external or served by the application server or the application itself.

Figure 4.1. Application Architecture

Figure 4.1, â&#x20AC;&#x153;Application Architectureâ&#x20AC;? above gives the basic architecture of an application made with the Vaadin framework, with all the major elements, which are introduced below and discussed in detail in this chapter. First of all, an application that uses Vaadin must define an application class that inherits the abstract com.vaadin.Application class. The application class must implement the init() method. public class MyApp extends com.vaadin.Application { public void init() { ... initialization code goes here ... } }

Besides acting as the entry-point in the servlet, the Application class provides facilities for window access, execution control, and theme selection. The application API may seem similar to Java Servlet API, but that is only superficial. Vaadin framework associates requests with sessions so that an application class instance is really a session object. Because of this, you can develop web applications much like you would develop desktop applications. The most important thing in the initialization is the creation of the main window (see below), which any application has. This, and the deployment of the application as a Java Servlet in the Servlet

44

Overview


Writing a Web Application

container, as described in Section 4.8, “Setting Up the Application Environment”, are the minimal requirements for an application. Below is a short overview of the basic elements of an application: Windows

An application always has a main window, as described in Section 4.2, “Managing the Main Window”. An application can actually have a number of such applicationlevel windows, all bound to the same application session, as described in Section 11.2, “Application-Level Windows”. Application-level windows can contain nonnative sub-windows, which are essentially floating layout components handled inside the browser.

User Interface Components

The user interface consists of UI components that are created and laid out by the application. User interaction with the components causes events (see below) related to the component, which the application must handle. Most components are bound to some data using the Data Model (see below). You can make your own UI components through either inheritance or composition. For a thorough reference of UI components, see Chapter 5, User Interface Components, for layout components, see Chapter 6, Managing Layout, and for composing components, see Section 5.21, “Component Composition with CustomComponent”.

Events and Listeners

Events, and listeners that handle events, are the basis of handling user interaction in an application. Section 3.5, “Events and Listeners” gave an introduction to events and listeners from an architectural point-ofview, while Section 4.4, “Handling Events with Listeners” later in this chapter takes a more practical view.

Resources

A user interface can display images or have links to web pages or downloadable documents. These are resources, which can be external or provided by the web server or the application itself. Section 4.5, “Referencing Resources” gives a practical overview of the different types of resources.

Themes

The presentation and logic of the user interface are separated. While the UI logic is handled as Java code, the presentation is defined in themes as CSS. Vaadin provides a default theme. User-defined themes can, in addition to style sheets, include HTML templates that define custom layouts and other theme resources, such as images.Themes are discussed in detail in Chapter 8, Themes, custom layouts in Section 6.13, “Custom Layouts”, and theme resources in Section 4.5.4, “Theme Resources”.

Data Binding

Field components are essentially views to data, represented in a data model. Using the data model, the components can update the application data directly, without

Overview

45


Writing a Web Application

the need for any control code. A field component model is always bound to a property, an item, or a container, depending on the field type. While all the components have a default data model, they can be bound to a userdefined data source. For example, you can bind a table component to an SQL query response. For a complete overview of data binding in Vaadin, please refer to Chapter 9, Binding Components to Data.

4.2. Managing the Main Window As explained in Section 11.1, “Special Characteristics of AJAX Applications”, an AJAX web application usually runs in a single "web page" in a browser window. The page is generally not reloaded after it is opened initially, but it communicates user interaction with the server through AJAX communications. A window in an AJAX application is therefore more like a window in a desktop application and less like a web page. A Window is the top-level container of a user interface displayed in a browser window. As an AJAX application typically runs on a single "page" (URL), there is usually just one window -- the main window. The main window can be accessed using the URL of the application. You set the main window with the setMainWindow() method of the Application class. import com.vaadin.ui.*; public class HelloWorld extends com.vaadin.Application { public void init() { Window main = new Window("The Main Window"); setMainWindow(main); ... fill the main window with components ... } }

You can add components to the main window, or to any other window, with the addComponent() method, which actually adds the given component to the root layout component bound to the window. If you wish to use other than the default root layout, you can set it with setContent(), as explained in Section 6.2, “Window and Panel Root Layout”. Vaadin has two basic kinds of windows: application-level windows, such as the main window, and child windows (or sub-windows) inside the application-level windows. The child windows are explained in the next section, while application-level windows are covered in Section 11.2, “Application-Level Windows”.

4.3. Child Windows An application-level window can have a number of floating child windows. They are managed by the client-side JavaScript runtime of Vaadin using HTML features. Vaadin allows opening and closing child windows, refreshing one window from another, resizing windows, and scrolling the window content. Child windows are typically used for Dialog Windows and Multiple Document Interface applications. Child windows are by default not modal; you can set them modal as described in Section 4.3.3, “Modal Windows”. As with all user interface components, the appearance of a window and its contents is defined with themes.

46

Managing the Main Window


Writing a Web Application

User control of a child window is limited to moving, resizing, and closing the window. Maximizing or minimizing are not yet supported.

4.3.1. Opening and Closing a Child Window You can open a new window by creating a new Window object and adding it to the main window with addWindow() method of the Application class. mywindow = new Window("My Window"); mainwindow.addWindow(mywindow);

You close the window in a similar fashion, by calling the removeWindow() of the Application class: myapplication.removeWindow (mywindow);

The user can, by default, close a child window by clicking the close button in the upper-right corner of the window. You can disable the button by setting the window as read-only with setReadOnly(true). Notice that you could disable the button also by making it invisible in CSS with a "display: none" formatting. The problem with such a cosmetic disabling is that a malicious user might re-enable the button and close the window, which might cause problems and possibly be a security hole. Setting the window as read-only not only disables the close button on the client side, but also prevents processing the close event on the server side. The following example demonstrates the use of a child window in an application. The example manages the window using a custom component that contains a button for opening and closing the window. /** Component contains a button that allows opening a window. */ public class WindowOpener extends CustomComponent implements Window.CloseListener { Window mainwindow; // Reference to main window Window mywindow; // The window to be opened Button openbutton; // Button for opening the window Button closebutton; // A button in the window Label explanation; // A descriptive text public WindowOpener(String label, Window main) { mainwindow = main; // The component contains a button that opens the window. final VerticalLayout layout = new VerticalLayout(); openbutton = new Button("Open Window", this, "openButtonClick"); explanation = new Label("Explanation"); layout.addComponent(openbutton); layout.addComponent(explanation); setCompositionRoot(layout); } /** Handle the clicks for the two buttons. */ public void openButtonClick(Button.ClickEvent event) { /* Create a new window. */ mywindow = new Window("My Dialog"); mywindow.setPositionX(200); mywindow.setPositionY(100); /* Add the window inside the main window. */ mainwindow.addWindow(mywindow);

Opening and Closing a Child Window

47


Writing a Web Application

/* Listen for close events for the window. */ mywindow.addListener(this); /* Add components in the window. */ mywindow.addComponent( new Label("A text label in the window.")); closebutton = new Button("Close", this, "closeButtonClick"); mywindow.addComponent(closebutton); /* Allow opening only one window at a time. */ openbutton.setEnabled(false); explanation.setValue("Window opened"); } /** Handle Close button click and close the window. */ public void closeButtonClick(Button.ClickEvent event) { /* Windows are managed by the application object. */ mainwindow.removeWindow(mywindow); /* Return to initial state. */ openbutton.setEnabled(true); explanation.setValue("Closed with button"); } /** In case the window is closed otherwise. */ public void windowClose(CloseEvent e) { /* Return to initial state. */ openbutton.setEnabled(true); explanation.setValue("Closed with window controls"); } }

The example implements a custom component that inherits the CustomComponent class. It consists of a Button that it uses to open a window and a Label to describe the state of the window. When the window is open, the button is disabled. When the window is closed, the button is enabled again. You can use the above custom component in the application class with: public void init() { Window main = new Window("The Main Window"); setMainWindow(main); addComponent(new WindowOpener("Window Opener", main)); }

When added to an application, the screen will look as illustrated in the following screenshot:

48

Opening and Closing a Child Window


Writing a Web Application

Figure 4.2. Opening a Child Window

4.3.2. Window Positioning When created, a window will have a default size and position.You can specify the size of a window with setHeight() and setWidth() methods. You can set the position of the window with setPositionX() and setPositionY() methods. /* Create a new window. */ mywindow = new Window("My Dialog"); /* Set window size. */ mywindow.setHeight("200px"); mywindow.setWidth("400px"); /* Set window position. */ mywindow.setPositionX(200); mywindow.setPositionY(50);

Notice that the size of the main window is unknown and the getHeight and getWidth methods will return -1.

4.3.3. Modal Windows A modal window is a child window that has to be closed by the user before the use of the parent window can continue. Dialog windows are typically modal. The advantage of modal windows is the simplification of user interaction, which may contribute to the clarity of the user interface. Modal windows are also easy to use from a development perspective, because as user interaction is isolated to them, changes in application state are more limited while the modal window is open. The disadvantage of modal windows is that they can restrict workflow too much.

Window Positioning

49


Writing a Web Application

Figure 4.3. Screenshot of the Modal Window Demo Application

Depending on theme settings, the parent window may be grayed while the modal window is open. The demo application of Vaadin includes an example of using modal windows. Figure 4.3, “Screenshot of the Modal Window Demo Application” above is from the demo application. The example includes the source code. Security Warning Modality of child windows is purely a client-side feature and can be circumvented with client-side attack code. You should not trust in the modality of child windows in security-critical situations such as login windows.

4.4. Handling Events with Listeners Let us put into practice what we learned of event handling in Section 3.5, “Events and Listeners”. You can handle events in three basic ways, as shown below. The following example follows a typical pattern where you have a Button component and a listener that handles user interaction (clicks) communicated to the application as events. Here we define a class that listens click events. public class TheButton implements Button.ClickListener { Button thebutton; /** Creates button into given container. */ public TheButton(AbstractComponentContainer container) { thebutton = new Button ("Do not push this button"); thebutton.addListener(this); container.addComponent(thebutton); } /** Handle button click events from the button. */ public void buttonClick (Button.ClickEvent event) { thebutton.setCaption ("Do not push this button again");

50

Handling Events with Listeners


Writing a Web Application

} }

As an application often receives events for several components of the same class, such as multiple buttons, it has to be able to distinguish between the individual components. There are several techniques to do this, but probably the easiest is to use the property of the received event, which is set to the object sending the event. This requires keeping at hand a reference to every object that emits events. public class TheButtons implements Button.ClickListener { Button thebutton; Button secondbutton; /** Creates two buttons in given container. */ public TheButtons(AbstractComponentContainer container) { thebutton = new Button ("Do not push this button"); thebutton.addListener(this); container.addComponent(thebutton); secondbutton = new Button ("I am a button too"); secondbutton.addListener(this); container.addComponent (secondbutton); } /** Handle button click events from the two buttons. */ public void buttonClick (Button.ClickEvent event) { if (event.getButton() == thebutton) thebutton.setCaption("Do not push this button again"); else if (event.getButton() == secondbutton) secondbutton.setCaption("I am not a number"); } }

Another solution to handling multiple events of the same class involves attaching an event source to a listener method instead of the class. An event can be attached to a method using another version of the addListener() method, which takes the event handler method as a parameter either as a name of the method name as a string or as a Method object. In the example below, we use the name of the method as a string. public class TheButtons2 { Button thebutton; Button secondbutton; /** Creates two buttons in given container. */ public TheButtons2(AbstractComponentContainer container) { thebutton = new Button ("Do not push this button"); thebutton.addListener(Button.ClickEvent.class, this, "theButtonClick"); container.addComponent(thebutton); secondbutton = new Button ("I am a button too"); secondbutton.addListener(Button.ClickEvent.class, this, "secondButtonClick"); container.addComponent (secondbutton); } public void theButtonClick (Button.ClickEvent event) { thebutton.setCaption ("Do not push this button again"); } public void secondButtonClick (Button.ClickEvent event) { secondbutton.setCaption ("I am not a number!"); } }

Handling Events with Listeners

51


Writing a Web Application

Adding a listener method with addListener() is really just a wrapper that creates a com.vaadin.event.ListenerMethod listener object, which is an adapter from a listener class to a method. It implements the java.util.EventListener interface and can therefore work for any event source using the interface. Notice that not all listener classes necessarily inherit the EventListener interface. The third way, which uses anonymous local class definitions, is often the easiest as it does not require cumbering the managing class with new interfaces or methods. The following example defines an anonymous class that inherits the Button.ClickListener interface and implements the buttonClick() method. public class TheButtons3 { Button thebutton; Button secondbutton; /** Creates two buttons in given container. */ public TheButtons3(AbstractComponentContainer container) { thebutton = new Button ("Do not push this button"); /* Define a listener in an anonymous class. */ thebutton.addListener(new Button.ClickListener() { /* Handle the click. */ public void buttonClick(ClickEvent event) { thebutton.setCaption ( "Do not push this button again"); } }); container.addComponent(thebutton); secondbutton = new Button ("I am a button too"); secondbutton.addListener(new Button.ClickListener() { public void buttonClick(ClickEvent event) { secondbutton.setCaption ("I am not a number!"); } }); container.addComponent (secondbutton); } }

Other techniques for separating between different sources also exist. They include using object properties, names, or captions to separate between them. Using captions or any other visible text is generally discouraged, as it may create problems for internationalization. Using other symbolic strings can also be dangerous, because the syntax of such strings is checked only runtime. Events are usually emitted by the framework, but applications may need to emit them too in some situations, such as when updating some part of the UI is required. Events can be emitted using the fireEvent(Component.Event) method of AbstractComponent.The event is then relayed to all the listeners of the particular event class for the object. Some components have a default event type, for example, a Button has a nested Button.ClickEvent class and a corresponding Button.ClickListener interface. These events can be triggered with fireComponentEvent().

4.5. Referencing Resources Web applications work over the web and have various resources, such as images or downloadable files, that the web browser has to get from the server. These resources are typically used in Embedded (images) or Link (downloadable files) user interface components.Various components, such as TabSheet, can also include icons, which are also handled as resources.

52

Referencing Resources


Writing a Web Application

A web server can handle many of such requests for static resources without having to ask them from the application, or the Application object can provide them. For dynamic resources, the user application must be able to create them dynamically. Vaadin provides resource request interfaces for applications so that they can return various kinds of resources, such as files or dynamically created resources. These include the StreamResource class and URI and parameter handlers described in Section 11.5.1, “URI Handlers” and Section 11.5.2, “Parameter Handlers”, respectively. Vaadin provides also low-level facilities for retrieving the URI and other parameters of a HTTP request. We will first look into how applications can provide various kinds of resources and then look into low-level interfaces for handling URIs and parameters to provide resources and functionalities. Notice that using URI or parameter handlers to create "pages" is not meaningful in Vaadin or in AJAX applications generally. Please see Section 11.1, “Special Characteristics of AJAX Applications” for a detailed explanation.

4.5.1. Resource Interfaces and Classes Vaadin has two interfaces for resources: a generic Resource interface and a more specific ApplicationResource interface for resources provided by the application.

Figure 4.4. Resource Interface and Class Diagram

ApplicationResource resources are managed by the Application class. When you create such a resource, you give the application object to the constructor. The constructor registers the resource in the application using the addResource method. Application manages requests for the resources and allows accessing resources using a URI. The URI consists of the base name of the application and a relative name of the resource. The relative name is "APP/"+resourceid+"/"+filename, for example "APP/1/myimage.png". The resourceid is a generated numeric identifier to make resources unique, and filename is the file name of the resource given in the constructor of its class. However, the application using a resource does not usually need to consider its URI. It only needs to give the resource to an appropriate Embedded or Link or some other user interface component, which manages the rendering of the URI.

4.5.2. File Resources File resources are files stored anywhere in the file system. The use of file resources generally falls into two main categories: downloadable files and embedded images. A file object that can be accessed as a file resource is defined with the standard java.io.File class. You can create the file either with an absolute or relative path, but the base path of the

Resource Interfaces and Classes

53


Writing a Web Application

relative path depends on the installation of the web server. For example, in Apache Tomcat, the default current directory is the installation path of Tomcat.

4.5.3. Class Loader Resources The ClassResource allows resources to be loaded from the deployed package of the application using Java Class Loader. The one-line example below loads an image resource from the application package and displays it in an Embedded component. mainwindow.addComponent(new Embedded ("", new ClassResource("smiley.jpg", mainwindow.getApplication())));

4.5.4. Theme Resources Theme resources are files included in a theme, typically images. See Chapter 8, Themes for more information on themes.

4.5.5. Stream Resources Stream resources are application resources that allow creating dynamic resource content. Charts are typical examples of dynamic images. To define a stream resource, you need to implement the StreamResource.StreamSource interface and its getStream method. The method needs to return an InputStream from which the stream can be read. The following example demonstrates the creation of a simple image in PNG image format. import java.awt.image.*; public class MyImageSource implements StreamResource.StreamSource { ByteArrayOutputStream imagebuffer = null; int reloads = 0; /* We need to implement this method that returns * the resource as a stream. */ public InputStream getStream () { /* Create an image and draw something on it. */ BufferedImage image = new BufferedImage (200, 200, BufferedImage.TYPE_INT_RGB); Graphics drawable = image.getGraphics(); drawable.setColor(Color.lightGray); drawable.fillRect(0,0,200,200); drawable.setColor(Color.yellow); drawable.fillOval(25,25,150,150); drawable.setColor(Color.blue); drawable.drawRect(0,0,199,199); drawable.setColor(Color.black); drawable.drawString("Reloads="+reloads, 75, 100); reloads++; try { /* Write the image to a buffer. */ imagebuffer = new ByteArrayOutputStream(); ImageIO.write(image, "png", imagebuffer); /* Return a stream from the buffer. */ return new ByteArrayInputStream( imagebuffer.toByteArray()); } catch (IOException e) { return null; }

54

Class Loader Resources


Writing a Web Application

} }

The content of the generated image is dynamic, as it updates the reloads counter with every call. The ImageIO.write() method writes the image to an output stream, while we had to return an input stream, so we stored the image contents to a temporary buffer. You can use resources in various ways. Some user interface components, such as Link and Embedded, take their parameters as a resource. Below we display the image with the Embedded component. The StreamResource constructor gets a reference to the application and registers itself in the application's resources. Assume that main is a reference to the main window and this is the application object. // Create an instance of our stream source. StreamResource.StreamSource imagesource = new MyImageSource (); // Create a resource that uses the stream source and give it a name. // The constructor will automatically register the resource in // the application. StreamResource imageresource = new StreamResource(imagesource, "myimage.png", this); // Create an embedded component that gets its contents // from the resource. main.addComponent(new Embedded("Image title", imageresource));

The image will look as follows:

Figure 4.5. Screenshot of the stream resource example with an embedded image

We named the resource as myimage.png. The application adds a resource key to the file name o f t h e r e s o u r c e t o m a ke i t u n i q u e. T h e f u l l U R I w i l l b e l i ke http://localhost:8080/testbench/APP/1/myimage.png. The end APP/1/myimage.png is the relative part of the URI.You can get the relative part of a resource's URI from the application with Application.getRelativeLocation(). Another solution for creating dynamic content is an URI handler, possibly together with a parameter handler. See Section 11.5.1, “URI Handlers” and Section 11.5.2, “Parameter Handlers”.

Stream Resources

55


Writing a Web Application

4.6. Shutting Down an Application A user can log out or close the web page or browser, so a session and the associated application instance can end. Ending an application can be initiated by the application logic. Otherwise, it will be ended automatically when the Servlet session times out.

4.6.1. Closing an Application If the user quits the application through the user interface, an event handler should call the close() method in the Application class to shutdown the session. In the following example, we have a Logout button, which ends the user session. Button closeButton = new Button("Logout"); closeButton.addListener(new Button.ClickListener() { @Override public void buttonClick(ClickEvent event) { getMainWindow().getApplication().close(); } }); main.addComponent(closeButton);

You will soon notice that closing the application simply reloads the application with a new Application instance. You can set the window to redirect to a different URL (that does not reload the application) with setLogoutURL. In your application class, write: setLogoutURL("/logout.html");

4.6.2. Handling the Closing of a Window Closing the main window (or all application-level windows) does not close session and the application instance will be left hanging. You need to program such behaviour by handling the close events of the windows. If the user closes a browser window, such as the main window or any other application-level window, the window will send a final AJAX request to the server, which will fire a Window.CloseEvent for the closed window.You can handle the event with a Window.CloseListener. In case the user closes the browser, the event is fired for every open window. // Close the application if the main window is closed. main.addListener(new Window.CloseListener(){ @Override public void windowClose(CloseEvent e) { System.out.println("Closing the application"); getMainWindow().getApplication().close(); } });

Notice that refreshing a window means closing and reopening it. Therefore, if you have a close handler as above, the user loses the possibility to refresh the browser window. In the likely case that the browser crashes, no close event is communicated to the server. As the server has no way of knowing about the problem, and the session will be left hanging until the session timeout expires. During this time, the user can restart the browser, open the application URL, and the main window will be rendered where the user left off. This can be desired behaviour in many cases, but sometimes it is not and can create a security problem.

56

Shutting Down an Application


Writing a Web Application

4.7. Handling Errors 4.7.1. Error Indicator and message All components have a built-in error indicator that can be set explicitly with setComponentError() or can be turned on implicitly if validating the component fails. As with component caption, the placement of the indicator is managed by the layout in which the component is contained. Usually, the error indicator is placed right of the caption text. Hovering the mouse pointer over the field displays the error message. The following example shows how you can set the component error explicitly. The example essentially validates field value without using an actual validator. // Create a field. final TextField textfield = new TextField("Enter code"); main.addComponent(textfield); // Let the component error be initially clear. textfield.setComponentError(null); // (actually the default) // Have a button right of the field (and align it properly). final Button button = new Button("Ok!"); main.addComponent(button); ((VerticalLayout)main.getLayout()) .setComponentAlignment(button, Alignment.BOTTOM_LEFT); // Handle button clicks button.addListener(new Button.ClickListener() { public void buttonClick(ClickEvent event) { // If the field value is bad, set its error. // (Allow only alphanumeric characters.) if (! ((String) textfield.getValue()).matches("^\\w*$")) { // Put the component in error state and // set the error message. textfield.setComponentError( new UserError("Must be letters and numbers")); } else { // Otherwise clear it. textfield.setComponentError(null); } } });

Figure 4.6. Error indicator active

The Form component handles and displays also the errors of its contained fields so that it displays both the error indicator and the message in a special error indicator area. See Section 5.17, “Form” and Section 5.17.3, “Validating Form Input” for details on the Form component and validation of form input.

4.7.2. Notifications Notifications are error or information boxes that appear typically at the center of the screen. A notification box has a caption and optional description and icon. The box stays on the screen

Handling Errors

57


Writing a Web Application

either for a defined time or until the user clicks it. The notification type defines the default appearance and behaviour of a notification. Notifications are always associated with a window object, which can be a child window (the positioning is always relative to the entire browser view). The Window class provides a showNotification() method for displaying notifications. The method takes the caption and an optional description and notification type as parameters. The method also accepts a notification object of type Window.Notification, as described further below. mainwindow.showNotification("This is the caption", "This is the description");

Figure 4.7. Notification

The caption and description are, by default, written on the same line. If you want to have a line break between them, use the XHTML line break markup "<br/>". You can use any XHTML markup in the caption and description of a notification. If it is possible to get the notification content from user input, you should sanitize the content carefully, as noted in Section 11.10.1, â&#x20AC;&#x153;Sanitizing User Input to Prevent Cross-Site Scriptingâ&#x20AC;?. main.showNotification("This is a warning", "<br/>This is the <i>last</i> warning", Window.Notification.TYPE_WARNING_MESSAGE);

Figure 4.8. Notification with Formatting

The notification type defines the overall default style and behaviour of a notification. If no notification type is given, the "humanized" type is used as the default. The notification types, listed below, are defined in the Window.Notification class.

58

TYPE_HUMANIZED_MESSAGE

A user-friendly message that does not annoy too much: it does not require confirmation by clicking and disappears quickly. It is centered and has a neutral gray color.

TYPE_WARNING_MESSAGE

Warnings are messages of medium importance. They are displayed with colors that are neither neutral nor too distractive. A warning is displayed for 1.5 seconds, but the user can click the message box to dismiss it. The user can continue to interact with the application while the warning is displayed.

Notifications


Writing a Web Application

TYPE_ERROR_MESSAGE

Error messages are notifications that require the highest user attention, with alert colors and by requiring the user to click the message to dismiss it. The error message box does not itself include an instruction to click the message, although the close box in the upper right corner indicates it visually. Unlike with other notifications, the user can not interact with the application while the error message is displayed.

TYPE_TRAY_NOTIFICATION

Tray notifications are displayed in the "system tray" area, that is, in the lower-right corner of the browser view. As they do not usually obsure any user interface, they are displayed longer than humanized or warning messages, 3 seconds by default.The user can continue to interact with the application normally while the tray notification is displayed.

All of the features of specific notification types can be controlled with the attributes of Window.Notification. You can pass an explicitly created notification object to the showNotification() method. // Create a notification with default settings for a warning. Window.Notification notif = new Window.Notification( "Be warned!", "This message lurks in the top-left corner!", Window.Notification.TYPE_WARNING_MESSAGE); // Set the position. notif.setPosition(Window.Notification.POSITION_TOP_LEFT); // Let it stay there until the user clicks it notif.setDelayMsec(-1); // Show it in the main window. main.showNotification(notif);

The setPosition() method allows setting the positioning of the notification. The method takes as its parameter any of the constants: Window.Notification.POSITION_CENTERED Window.Notification.POSITION_CENTERED_TOP Window.Notification.POSITION_CENTERED_BOTTOM Window.Notification.POSITION_TOP_LEFT Window.Notification.POSITION_TOP_RIGHT Window.Notification.POSITION_BOTTOM_LEFT Window.Notification.POSITION_BOTTOM_RIGHT The setDelayMSec() allows you to set the time in milliseconds for how long the notification is displayed. Parameter value -1 means that the message is displayed until the user clicks the message box. It also prevents interaction with other parts of the application window, as is default behaviour for error messages. It does not, however, add a close box that the error notification has.

Notifications

59


Writing a Web Application

4.7.3. Handling Uncaught Exceptions Application development with Vaadin follows the event-driven programming model. Mouse and keyboard events in the client cause (usually higher-level) events on the server-side, which can be handled with listeners, and that is how most of the application logic works. Handling the events can result in exceptions either in the application logic or in the framework itself, but some of them may not be caught properly. For example, in the following code excerpt, we throw an error in an event listener but do not catch it, so it falls to the framework. final Button button = new Button ("Fail Me"); button.addListener(new Button.ClickListener() { public void buttonClick(ClickEvent event) { // Throw some exception. throw new RuntimeException("You can't catch this."); } });

Any such exceptions that occur in the call chain, but are not caught at any other level, are eventually caught by the terminal adapter in ApplicationServlet, the lowest-level component that receives client requests. The terminal adapter passes all such caught exceptions as events to the error listener of the Application instance through the Terminal.ErrorListener interface. The Application class does not, by default, throw such exceptions forward. The reason for this error-handling logic lies in the logic that handles component state synchronization between the client and the server. We want to handle all the serialized variable changes in the client request, because otherwise the client-side and server-side component states would become unsynchronized very easily, which could put the entire application in an invalid state. The default implementation of the Terminal.ErrorListener interface in the Application class simply prints the error to console. It also tries to find out a component related to the error. If the exception occurred in a listener attached to a component, that component is considered as the component related to the exception. If a related component is found, the error handler sets the component error for it, the same attribute which you can set with setComponentError(). In UI, the component error is shown with a small red "!" -sign (in the default theme). If you hover the mouse pointer over it, you will see the entire backtrace of the exception in a large tooltip box, as illustrated in Figure 4.9, â&#x20AC;&#x153;Uncaught Exception in Component Error Indicatorâ&#x20AC;? for the above code example.

Figure 4.9. Uncaught Exception in Component Error Indicator

60

Handling Uncaught Exceptions


Writing a Web Application

You can change the logic of handling the terminal errors easily by overriding the terminalError() method in your application class (the one that inherits Application) or by setting a custom error listener with the setErrorHandler method. You can safely discard the default handling or extend its usage with your custom error handling or logging system. In the example code below, the exceptions are also reported as notifications in the main window. @Override public void terminalError(Terminal.ErrorEvent event) { // Call the default implementation. super.terminalError(event); // Some custom behaviour. if (getMainWindow() != null) { getMainWindow().showNotification( "An unchecked exception occured!", event.getThrowable().toString(), Notification.TYPE_ERROR_MESSAGE); } }

Handling other exceptions works in the usual way for Java Servlets. Uncaught exceptions are finally caught and handled by the application server.

4.8. Setting Up the Application Environment While more and more server based frameworks, libraries, standards, and architectures for Java are invented to make the programmer's life easier, software deployment seems to get harder and harder. For example, Java Enterprise Beans tried to make the creation of persistent and networked objects easy and somewhat automatic, but the number of deployment descriptions got enormous. As Vaadin lives in a Java Servlet container, it must follow the rules, but it tries to avoid adding extra complexity. All Vaadin applications are deployed as Java web applications, which can be packaged as WAR files. For a detailed tutorial on how web applications are packaged, please refer to any Java book that discusses Servlets. Sun has an excellent reference online at http://java.sun.com/j2ee/tutorial/1_3-fcs/doc/WCC3.html [http://java.sun.com/j2ee/tutorial/1_3-fcs/doc/WCC3.html].

4.8.1. Creating Deployable WAR in Eclipse To deploy an application to a web server, you need to create a WAR package. Here we give the instructions for Eclipse. Open project properties and first set the name and destination of the WAR file in Tomcat Export to WAR settings tab. Exporting to WAR is done by selecting Export to WAR from Tomcat Project in project context menu (just click calc with the right mouse button on Package contents tree).

4.8.2. Web Application Contents The following files are required in a web application in order to run it.

Web application organization WEB-INF/web.xml

This is the standard web application descriptor that defines how the application is organized. You can refer

Setting Up the Application Environment

61


Writing a Web Application

to any Java book about the contents of this file. Also see an example in Example 4.1, â&#x20AC;&#x153;web.xmlâ&#x20AC;?. WEB-INF/lib/vaadin-6.2.0.jar This is the Vaadin library. It is included in the product package in lib directory. Your application classes

You must include your application classes either in a JAR file in WEB-INF/lib or as classes in WEB-INF/classes

Your own theme files (OPTIONAL)

If your application uses a special theme (look and feel), yo u mu s t i n c l u d e i t i n WEB-INF/lib/themes/themename directory.

4.8.3. Deployment Descriptor web.xml The deployment descriptor is an XML file with the name web.xml in the WEB-INF directory of a web application. It is a standard component in Java EE describing how a web application should be deployed. The structure of the deployment descriptor is illustrated by the following example. You simply deploy applications as servlets implemented by the special com.vaadin.terminal.gwt.server.ApplicationServlet wrapper class. The class of the actual application is specified by giving the application parameter with the name of the specific application class to the servlet. The servlet is then connected to a URL in a standard way for Java Servlets.

Example 4.1. web.xml <?xml version="1.0" encoding="UTF-8"?> <web-app id="WebApp_ID" version="2.4" xmlns="http://java.sun.com/xml/ns/j2ee" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://java.sun.com/xml/ns/j2ee http://java.sun.com/xml/ns/j2ee/web-app_2_4.xsd"> <servlet> <servlet-name>myservlet</servlet-name> <servlet-class> com.vaadin.terminal.gwt.server.ApplicationServlet </servlet-class> <init-param> <param-name>application</param-name> <param-value>MyApplicationClass</param-value> </init-param> </servlet> <servlet-mapping> <servlet-name>myservlet</servlet-name> <url-pattern>/*</url-pattern> </servlet-mapping> </web-app>

The descriptor defines a servlet with name myservlet. The servlet class, com.vaadin.terminal.gwt.server.ApplicationServlet, is provided by Vaadin framework and it should be the same for all Vaadin projects. The servlet takes the class name Calc of the user application class as a parameter, including the full package path to the class. If the class is in the default package the package path is obviously not used.

62

Deployment Descriptor web.xml


Writing a Web Application

The url-pattern is defined above as /*. This matches to any URL under the project context. We defined above the project context as myproject so the application URL will be http://localhost:8080/myproject/. If the project were to have multiple applications or servlets, they would have to be given different names to distinguish them. For example, url-pattern /myapp/* would match a URL such as http://localhost:8080/myproject/myapp/. Notice that the slash and the asterisk must be included at the end of the pattern. Notice also that if the URL pattern is other than root /* (such as /myapp/*), you will also need to make a servlet mapping to /VAADIN/* (unless you are serving it statically as noted below). For example: ... <servlet-mapping> <servlet-name>myservlet</servlet-name> <url-pattern>/myurl/*</url-pattern> </servlet-mapping> <servlet-mapping> <servlet-name>myservlet</servlet-name> <url-pattern>/VAADIN/*</url-pattern> </servlet-mapping>

You do not have to provide the above /VAADIN/* mapping if you serve both the widget sets and (custom and default) themes statically in WebContent/VAADIN/ directory. The mapping simply allows serving them dynamically from the Vaadin JAR. Serving them statically is recommended for production environments as it is much faster. For a complete example on how to deploy applications, see the demos included in the Vaadin installation package, especially the WebContent/WEB-INF directory.

Deployment Descriptor Parameters Deployment descriptor can have many parameters and options that control the execution of a servlet. You can find a complete documentation of the deployment descriptor in Java Servlet Specification at http://java.sun.com/products/servlet/. By default, Vaadin applications run in debug mode, which should be used during development. This enables various debugging features. For production use, you should have put in your web.xml the following parameter: <context-param> <param-name>productionMode</param-name> <param-value>true</param-value> <description>Vaadin production mode</description> </context-param>

The parameter and the debug and production modes are described in detail in Section 11.4, â&#x20AC;&#x153;Debug and Production Modeâ&#x20AC;?. One often needed option is the session timeout. Different servlet containers use varying defaults for timeouts, such as 30 minutes for Apache Tomcat. You can set the timeout with: <session-config> <session-timeout>30</session-timeout> </session-config>

After the timeout expires, the close() method of the Application class will be called.You should implement it if you wish to handle timeout situations.

Deployment Descriptor web.xml

63


64


Chapter 5

User Interface Components 5.1. Overview .................................................................................................. 66 5.2. Interfaces and Abstractions ..................................................................... 67 5.3. Common Component Features ............................................................... 72 5.4. Label ....................................................................................................... 82 5.5. Link ......................................................................................................... 85 5.6. TextField ................................................................................................. 86 5.7. RichTextArea .......................................................................................... 87 5.8. Date and Time Input with DateField ........................................................ 88 5.9. Button ..................................................................................................... 93 5.10. CheckBox ............................................................................................. 94 5.11. Selecting Items ...................................................................................... 95 5.12. Table .................................................................................................... 106 5.13. Tree ..................................................................................................... 123 5.14. MenuBar ............................................................................................. 125 5.15. Embedded .......................................................................................... 126 5.16. Upload ................................................................................................ 128 5.17. Form .................................................................................................... 131 5.18. ProgressIndicator .............................................................................. 139 5.19. Slider ................................................................................................... 140 5.20. LoginForm .......................................................................................... 142

Book of Vaadin

65


User Interface Components

5.21. Component Composition with CustomComponent ........................... 145 This chapter provides an overview and a detailed description of all non-layout components in Vaadin.

5.1. Overview Vaadin provides a comprehensive set of user interface components and allows you to define custom components. Figure 5.1, â&#x20AC;&#x153;UI Component Inheritance Diagramâ&#x20AC;? illustrates the inheritance hierarchy of the UI component classes and interfaces. Interfaces are displayed in gray, abstract classes in orange, and regular classes in blue. An annotated version of the diagram is featured in the Vaadin Cheat Sheet.

Figure 5.1. UI Component Inheritance Diagram

At the top of the interface hierarchy, we have the Component interface. At the top of the class hierarchy, we have the AbstractComponent class. It is inherited by two other abstract classes: AbstractField, inherited further by field components, and AbstractComponentContainer, inherited by various container and layout components. Components that are not bound to a content data model, such as labels and links, inherit AbstractComponent directly.

66

Overview


User Interface Components

The layout of the various components in a window is controlled, logically, by layout components, just like in conventional Java UI toolkits for desktop applications. In addition, with the CustomLayout component, you can write a custom layout as an XHTML template that includes the locations of any contained components. Looking at the inheritance diagram, we can see that layout components inherit the AbstractComponentContainer and the Layout interface. Layout components are described in detail in Chapter 6, Managing Layout. Looking at it from the perspective of an object hierarchy, we would have a Window object, which contains a hierachy of layout components, which again contain other layout components, field components, and other visible components. You can browse the built-in UI components of Vaadin library in the Sampler application of the Vaadin Demo. The Sampler shows a description, JavaDoc documentation, and a code samples for each of the components. In addition to the built-in components, many components are available as add-ons, either from the Vaadin Directory or from independent sources. Both commercial and free components exist. The installation of add-ons is described in Section 11.14, â&#x20AC;&#x153;Using Add-on Componentsâ&#x20AC;?.

5.2. Interfaces and Abstractions Vaadin user interface components are built on a skeleton of interfaces and abstract classes that define and implement the features common to all components and the basic logic how the component states are serialized between the server and the client. This section gives details on the basic component interfaces and abstractions. The layout and other component container abstractions are described in Chapter 6, Managing Layout. The interfaces that define the Vaadin data model are described in Chapter 9, Binding Components to Data.

Figure 5.2. Component Interfaces and Abstractions

Interfaces and Abstractions

67


User Interface Components

All components also implement the Paintable interface, which is used for serializing ("painting") the components to the client, and the reverse VariableOwner interface, which is needed for deserializing component state or user interaction from the client. In addition to the interfaces defined within the Vaadin framework, all components implement the java.io.Serializable interface to allow serialization. Serialization is needed in many clustering and cloud computing solutions.

5.2.1. Component Interface The Component interface is paired with the AbstractComponent class, which implements all the methods defined in the interface.

Component Tree Management Components are laid out in the user interface hierarchically. The layout is managed by layout components, or more generally components that implement the ComponentContainer interface. Such a container is the parent of the contained components. The getParent() method allows retrieving the parent component of a component. While there is a setParent(), you rarely need it as you usually add components with the addComponent() method of the ComponentContainer interface, which automatically sets the parent. A component does not know its parent when the component is created, so you can not refer to the parent in the constructor with getParent(). Also, it is not possible to fetch a reference to the application object with getApplication() before having a parent. For example, the following is invalid: public class AttachExample extends CustomComponent { public AttachExample() { // ERROR: We can't access the application object yet. ClassResource r = new ClassResource("smiley.jpg", getApplication()); Embedded image = new Embedded("Image:", r); setCompositionRoot(image); } }

Adding a component to an application triggers calling the attach() method for the component. Correspondingly, removing a component from a container triggers calling the detach() method. If the parent of an added component is already connected to the application, the attach() is called immediately from setParent(). public class AttachExample extends CustomComponent { public AttachExample() { } @Override public void attach() { super.attach(); // Must call. // Now we know who ultimately owns us. ClassResource r = new ClassResource("smiley.jpg", getApplication()); Embedded image = new Embedded("Image:", r); setCompositionRoot(image); } }

68

Component Interface


User Interface Components

The attachment logic is implemented in AbstractComponent, as described in Section 5.2.2, “AbstractComponent”.

5.2.2. AbstractComponent AbstractComponent is the base class for all user interface components. It is the (only) implementation of the Component interface, implementing all the methods defined in the interface. AbstractComponent has a single abstract method, getTag(), which returns the serialization identifier of a particular component class. It needs to be implemented when (and only when) creating entirely new components. AbstractComponent manages much of the serialization of component states between the client and the server. Creation of new components and serialization is described in Chapter 10, Developing Custom Components, and the server-side serialization API in Appendix A, User Interface Definition Language (UIDL).

5.2.3. Field Components (Field and AbstractField) Fields are components that have a value that the user can change through the user interface. Figure 5.3, “Field Components” illustrates the inheritance relationships and the important interfaces and base classes.

AbstractComponent

69


User Interface Components

Figure 5.3. Field Components

Field components are built upon the framework defined in the Field interface and the AbstractField base class. Fields are strongly coupled with the Vaadin data model. The field value is handled as a Property of the field component. Selection fields allow management of the selectable items through the Container interface. The description of the field interfaces and base classes is broken down in the following sections.

Field Interface The Field interface inherits the Component superinterface and also the Property interface to have a value for the field. AbstractField is the only class implementing the Field interface directly. The relationships are illustrated in Figure 5.4, â&#x20AC;&#x153;Field Interface Inheritance Diagramâ&#x20AC;?.

70

Field Components (Field and AbstractField)


User Interface Components

Figure 5.4. Field Interface Inheritance Diagram

You can set the field value with the setValue() and read with the getValue() method defined in the Property interface. The actual value type depends on the component. The Field interface defines a number of attributes, which you can retrieve or manipulate with the corresponding setters and getters. description

All fields have a description. Notice that while this attribute is defined in the Field component, it is implemented in AbstractField, which does not directly implement Field, but only through the AbstractField class.

required

A field can be marked as required and a required indicator marker (usually * character) is displayed in front of the field. If such fields are validated but are empty, the error indicator is shown and the component error is set to the text defined with the requiredError attribute (see below). Without validation, the required indicator is merely a visual guide.

requiredError

Defines the error message to show when a value is required for a field but no value is given. The error message is set as the component error for the field and is usually displayed in a tooltip. The Form component can display the error message in a special error indicator area.

Handling Field Value Changes Field inherits Property.ValueChangeListener to allow listening for field value changes and Property.Editor to allow editing values. When the value of a field changes, a Property.ValueChangeEvent is triggered for the field. You should not implement the valueChange() method in a class inheriting AbstractField, as it is already implemented in AbstractField. You should instead implement the method explicitly by adding the implementing object as a listener.

Field Components (Field and AbstractField)

71


User Interface Components

AbstractField Base Class AbstractField is the base class for all field components. In addition to the component features inherited from AbstractComponent, it implements a number of features defined in Property, Buffered, Validatable, and Component.Focusable interfaces.

5.3. Common Component Features The component base classes and interfaces provide a large number of features. Let us look at some of the most commonly needed features. Features not documented here can be found from the Java API Reference. The interface defines a number of properties, which you can retrieve or manipulate with the corresponding setters and getters.

5.3.1. Caption A caption is an explanatory textual label accompanying a user interface component, usually shown above, left of, or inside the component. The contents of a caption are automatically quoted, so no raw XHTML can be rendered in a caption. The caption can usually be given as the first parameter of a constructor or later with the setCaption() method. // New text field with caption "Name" TextField name = new TextField("Name"); layout.addComponent(name);

The caption of a component is, by default, managed and displayed by the layout component or component container in which the component is placed. For example, the VerticalLayout component shows the captions left-aligned above the contained components, while the FormLayout component shows the captions on the left side of the vertically laid components, with the captions and their associated components left-aligned in their own columns. The CustomComponent does not manage the caption of its composition root, so if the root component has a caption, it will not be rendered.

Figure 5.5. Caption Management by VerticalLayout and FormLayout components.

Some components, such as Button and Panel, manage the caption themselves and display it inside the component. Icon (see Section 5.3.4, â&#x20AC;&#x153;Iconâ&#x20AC;?) is closely related to caption and is usually displayed horizontally before or after it, depending on the component and the containing layout.

72

Common Component Features


User Interface Components

An alternative way to implement a caption is to use another component as the caption, typically a Label, a TextField, or a Panel. A Label, for example, allows highlighting a shortcut key with XHTML markup or to bind the caption to a data source. The Panel provides an easy way to add both a caption and a border around a component. Setting the caption with setCaption() will cause updating the component. A reimplementation of setCaption() should call the superclass implementation.

CSS Style Rules A caption will be rendered inside an HTML element that has the <component>-caption CSS style class, where <component> is the identifier of the component to which the caption is attached. For example, the caption of a TextField would have v-textfield-caption style.The containing layout may enclose a caption inside other caption-related elements.

5.3.2. Description and Tooltips All components (that inherit AbstractComponent) have a description separate from their caption. The description is usually shown as a tooltip that appears when the mouse pointer hovers over the component for a short time. You can set the description with setDescription() and retrieve with getDescription(). Button button = new Button("A Button"); button.setDescription("This is the tooltip");

The tooltip is shown in Figure 5.6, “Component Description as a Tooltip”.

Figure 5.6. Component Description as a Tooltip

A description is rendered as a tooltip in most components. Form shows it as text in the top area of the component, as described in Section 5.17.1, “Form as a User Interface Component”. When a component error has been set with setComponentError(), the error is usually also displayed in the tooltip, below the description (Form displays it in the bottom area of the form). Components that are in error state will also display the error indicator. See Section 4.7.1, “Error Indicator and message”. The description is actually not plain text, but you can use XHTML tags to format it. Such a rich text description can contain any HTML elements, including images. button.setDescription( "<h2><img src=\"../VAADIN/themes/sampler/icons/comment_yellow.gif\"/>"+ "A richtext tooltip</h2>"+ "<ul>"+ " <li>Use rich formatting with XHTML</li>"+ " <li>Include images from themes</li>"+ " <li>etc.</li>"+ "</ul>");

The result is shown in Figure 5.7, “A Rich Text Tooltip”.

Description and Tooltips

73


User Interface Components

Figure 5.7. A Rich Text Tooltip

Notice that the setter and getter are defined for all fields in the Field interface, not for all components in the Component interface.

5.3.3. Enabled The enabled property controls whether the user can actually use the component. A disabled component is visible, but grayed to indicate the disabled state. Components are always enabled by default. You can disable a component with setEnabled(false). Button enabled = new Button("Enabled"); enabled.setEnabled(true); // The default layout.addComponent(enabled); Button disabled = new Button("Disabled"); disabled.setEnabled(false); layout.addComponent(disabled);

Figure 5.8, “An Enabled and Disabled Button” shows the enabled and disabled buttons.

Figure 5.8. An Enabled and Disabled Button

A disabled component is automatically put in read-only state. No client interaction with such a component is sent to the server and, as an important security feature, the server-side component does not receive state updates from the client.

CSS Style Rules Disabled components have the v-disabled CSS style in addition to the component-specific style. To match a component with both the styles, you have to join the style class names with a dot as done in the example below. .v-textfield.v-disabled { border: dotted; }

This would make the border of all disabled text fields dotted. TextField disabled = new TextField("Disabled"); disabled.setValue("Read-only value"); disabled.setEnabled(false); layout.addComponent(disabled);

The result is illustrated in Figure 5.9, “Styling Disabled Components”.

74

Enabled


User Interface Components

Figure 5.9. Styling Disabled Components

5.3.4. Icon An icon is an explanatory graphical label accompanying a user interface component, usually shown above, left of, or inside the component. Icon is closely related to caption (see Section 5.3.1, â&#x20AC;&#x153;Captionâ&#x20AC;?) and is usually displayed horizontally before or after it, depending on the component and the containing layout. The icon of a component can be set with the setIcon() method. The image is provided as a resource, perhaps most typically a ThemeResource. // Component with an icon from a custom theme TextField name = new TextField("Name"); name.setIcon(new ThemeResource("icons/user.png")); layout.addComponent(name); // Component with an icon from another theme ('runo') Button ok = new Button("OK"); ok.setIcon(new ThemeResource("../runo/icons/16/ok.png")); layout.addComponent(ok);

The icon of a component is, by default, managed and displayed by the layout component or component container in which the component is placed. For example, the VerticalLayout component shows the icons left-aligned above the contained components, while the FormLayout component shows the icons on the left side of the vertically laid components, with the icons and their associated components left-aligned in their own columns. The CustomComponent does not manage the icon of its composition root, so if the root component has an icon, it will not be rendered.

Figure 5.10. Displaying an Icon from a Theme Resource.

Some components, such as Button and Panel, manage the icon themselves and display it inside the component.

CSS Style Rules An icon will be rendered inside an HTML element that has the v-icon CSS style class. The containing layout may enclose an icon and a caption inside elements related to the caption, such as v-caption.

5.3.5. Locale The locale property defines the country and language used in a component. You can use the locale information in conjunction with an internationalization scheme to acquire localized resources. Some components, such as DateField, use the locale for component localization. You can set the locale of a component (or the application) with setLocale(). // Component for which the locale is meaningful InlineDateField date = new InlineDateField("Datum");

Icon

75


User Interface Components

// German language specified with ISO 639-1 language // code and ISO 3166-1 alpha-2 country code. date.setLocale(new Locale("de", "DE")); date.setResolution(DateField.RESOLUTION_DAY); layout.addComponent(date);

The resulting date field is shown in Figure 5.11, â&#x20AC;&#x153;Set Locale for InlineDateFieldâ&#x20AC;?.

Figure 5.11. Set Locale for InlineDateField

You can get the locale of a component with getLocale(). If the locale is undefined for a component, that is, not explicitly set, the locale of the parent component is used. If none of the parent components have a locale set, the locale of the application is used, and if that is not set, the default system locale is set, as given by Locale.getDefault(). Because of the requirement that the component must be attached to the application, it is awkward to use getLocale() for internationalization. You can not use it in the constructor, so you would have to get the locale in attach() as shown in the following example: Button cancel = new Button() { @Override public void attach() { ResourceBundle bundle = ResourceBundle.getBundle( MyAppCaptions.class.getName(), getLocale()); setCaption(bundle.getString("CancelKey")); } }; layout.addComponent(cancel);

It is normally a better practice to get the locale from an application-global parameter and use it to get the localized resource right when the component is created. // Captions are stored in MyAppCaptions resource bundle // and the application object is known in this context. ResourceBundle bundle = ResourceBundle.getBundle(MyAppCaptions.class.getName(), getApplication().getLocale()); // Get a localized resource from the bundle Button cancel = new Button(bundle.getString("CancelKey")); layout.addComponent(cancel);

Selecting a Locale A common task in many applications is selecting a locale. This is done in the following example with a Select component. // The locale in which we want to have the language // selection list Locale displayLocale = Locale.ENGLISH; // All known locales

76

Locale


User Interface Components

final Locale[] locales = Locale.getAvailableLocales(); // Allow selecting a language. We are in a constructor of a // CustomComponent, so preselecting the current // language of the application can not be done before // this (and the selection) component are attached to // the application. final Select select = new Select("Select a language") { @Override public void attach() { setValue(getLocale()); } }; for (int i=0; i<locales.length; i++) { select.addItem(locales[i]); select.setItemCaption(locales[i], locales[i].getDisplayName(displayLocale)); // Automatically select the current locale if (locales[i].equals(getLocale())) select.setValue(locales[i]); } layout.addComponent(select); // Locale code of the selected locale final Label localeCode = new Label(""); layout.addComponent(localeCode); // A date field which language the selection will change final InlineDateField date = new InlineDateField("Calendar in the selected language"); date.setResolution(DateField.RESOLUTION_DAY); layout.addComponent(date); // Handle language selection select.addListener(new Property.ValueChangeListener() { public void valueChange(ValueChangeEvent event) { Locale locale = (Locale) select.getValue(); date.setLocale(locale); localeCode.setValue("Locale code: " + locale.getLanguage() + "_" + locale.getCountry()); } }); select.setImmediate(true);

The user interface is shown in Figure 5.12, â&#x20AC;&#x153;Selecting a Localeâ&#x20AC;?.

Figure 5.12. Selecting a Locale

Locale

77


User Interface Components

5.3.6. Read-Only The property defines whether the user can change the value of a component. As only Field components have a value that can be input or edited by the user, this method is mainly applicable to field components. The read-only property does not prevent changing the value of a component programmatically. TextField readwrite = new TextField("Read-Write"); readwrite.setValue("You can change this"); readwrite.setReadOnly(false); // The default layout.addComponent(readwrite); TextField readonly = new TextField("Read-Only"); readonly.setValue("You can't touch this!"); readonly.setReadOnly(true); layout.addComponent(readonly);

The resulting read-only text field is shown in Figure 5.13, â&#x20AC;&#x153;A Read-Only Component.â&#x20AC;?.

Figure 5.13. A Read-Only Component.

Even when a component is read-only, the user may be able to interact with some component features, such as scrolling. Client-side state modifications will not be communicated to the server-side and, more important, server-side field components will not accept changes to the value of a read-only component. This is an important security feature, because a malicious user can not fabricate state changes in a read-only component. Notice that this occurs at the level of AbstractField in setValue() and therefore only applies to the property value of the component; a component could itself accept some variable changes from the client-side.

CSS Style Rules The appearance of the component can change to indicate that the value is not editable. A readonly component will have the v-readonly style. The following CSS rule would make the text in all read-only TextField components appear in italic. .v-textfield.v-readonly { font-style: italic; }

An icon will be rendered inside an HTML element that has the v-icon CSS style class. The containing layout may enclose an icon and a caption inside elements related to the caption, such as v-caption.

5.3.7. Style Name The style name property defines one or more custom CSS style class names for the component. The getStyleName() returns the current style names as a space-separated list. The setStyleName() replaces all the styles with the given style name or a space-separated list of style names. You can also add and remove individual style names with addStylename() and removeStyleName(). A style name must be a valid CSS style name.

78

Read-Only


User Interface Components

Label label = new Label("This text has a lot of style"); label.addStyleName("mystyle"); layout.addComponent(label);

The style name will appear in the component's HTML element in two forms: literally as given and prefixed with the component class specific style name. For example, if you add a style name mystyle to a Button, the component would get both mystyle and v-button-mystyle styles. Neither form may conflict with built-in style names of Vaadin or GWT. For example, focus style would conflict with a built-in style of the same name, and an option style for a Select component would conflict with the built-in v-select-option style. The following CSS rule would apply the style to any component that has the mystyle style. .mystyle { font-family: font-style: font-size: font-weight: line-height: }

fantasy; italic; 25px; bolder; 30px;

The resulting styled component is shown in Figure 5.14, “Component with a Custom Style”

Figure 5.14. Component with a Custom Style

5.3.8. Visible Components can be hidden by setting the visible property to false. Also the caption, icon and any other component features are made hidden. Hidden components are not just invisible, but their content is not communicated to the browser at all. That is, they are not made invisible cosmetically with only CSS rules. This feature is important for security if you have components that contain security-critical information that must only be shown in specific application states. TextField readonly = new TextField("Read-Only"); readonly.setValue("You can't see this!"); readonly.setVisible(false); layout.addComponent(readonly);

The resulting invisible component is shown in Figure 5.15, “An Invisible Component.”.

Figure 5.15. An Invisible Component.

If you need to make a component only cosmetically invisible, you should use a custom theme to set it display: none style. This is mainly useful for certain special components such as ProgressIndicator, which have effects even when made invisible in CSS. If the hidden component has undefined size and is enclosed in a layout that also has undefined size, the containing layout will collapse when the component disappears. If you want to have the component keep its size, you have to make it invisible by setting all its font and other attributes to be transparent. In such cases, the invisible content of the component can be made visible easily in the browser. An invisible component has no particular CSS style class to indicate that it is hidden. The element does exist though, but has display: none style, which overrides any CSS styling.

Visible

79


User Interface Components

5.3.9. Sizing Components Vaadin components are sizeable; not in the sense that they were fairly large or that the number of the components and their features are sizeable, but in the sense that you can make them fairly large on the screen if you like, or small or whatever size. The Sizeable interface, shared by all components, provides a number of manipulation methods and constants for setting the height and width of a component in absolute or relative units, or for leaving the size undefined. The size of a component can be set with setWidth() and setHeight() methods.The methods take the size as a floating-point value. You need to give the unit of the measure as the second parameter for the above methods. The available units are listed in Table 5.1, “Size Units” below. mycomponent.setWidth(100, Sizeable.UNITS_PERCENTAGE); mycomponent.setWidth(400, Sizeable.UNITS_PIXELS);

Alternatively, you can speficy the size as a string. The format of such a string must follow the HTML/CSS standards for specifying measures. mycomponent.setWidth("100%"); mycomponent.setHeight("400px");

The "100%" percentage value makes the component take all available size in the particular direction (see the description of Sizeable.UNITS_PERCENTAGE in the table below). You can also use the shorthand method setSizeFull() to set the size to 100% in both directions. The size can be undefined in either or both dimensions, which means that the component will take the minimum necessary space. Most components have undefined size by default, but some layouts have full size in horizontal direction. You can set the height or width as undefined with Sizeable.SIZE_UNDEFINED parameter for setWidth() and setHeight(). You always need to keep in mind that a layout with undefined size may not contain components with defined relative size, such as "full size". See Section 6.12.1, “Layout Size” for details. The Table 5.1, “Size Units” lists the available units and their codes defined in the Sizeable interface.

80

Sizing Components


User Interface Components

Table 5.1. Size Units UNITS_PIXELS

px

The pixel is the basic hardware-specific measure of one physical display pixel.

UNITS_POINTS

pt

The point is a typographical unit, which is usually defined as 1/72 inches or about 0.35 mm. However, on displays the size can vary significantly depending on display metrics.

UNITS_PICAS

pc

The pica is a typographical unit, defined as 12 points, or 1/7 inches or about 4.233 mm. On displays, the size can vary depending on display metrics.

UNITS_EM

em A unit relative to the used font, the width of the upper-case "M" letter.

UNITS_EX

ex

UNITS_MM

mm A physical length unit, millimeters on the surface of a display device. However, the actual size depends on the display, its metrics in the operating system, and the browser.

UNITS_CM

cm A physical length unit, centimeters on the surface of a display device. However, the actual size depends on the display, its metrics in the operating system, and the browser.

UNITS_INCH

in

UNITS_PERCENTAGE %

A unit relative to the used font, the height of the lower-case "x" letter.

A physical length unit, inches on the surface of a display device. However, the actual size depends on the display, its metrics in the operating system, and the browser. A relative percentage of the available size. For example, for the toplevel layout 100% would be the full width or height of the browser window. The percentage value must be between 0 and 100.

If a component inside HorizontalLayout or VerticalLayout has full size in the namesake direction of the layout, the component will expand to take all available space not needed by the other components. See Section 6.12.1, â&#x20AC;&#x153;Layout Sizeâ&#x20AC;? for details.

5.3.10. Managing Input Focus When the user clicks on a component, the component gets the input focus, which is indicated by highlighting according to style definitions. If the component allows inputting text, the focus and insertion point are indicated by a cursor. Pressing the Tab key moves the focus to the component next in the focus order. Focusing is supported by all Field components and also by Form and Upload. The focus order or tab index of a component is defined as a positive integer value, which you can set with setTabIndex() and get with getTabIndex(). The tab index is managed in the context of the application-level Window in which the components are contained. The focus order can therefore jump between two any lower-level component containers, such as sub-windows or panels. The default focus order is determined by the natural hierarchical order of components in the order in which they were added under their parents. The default tab index is 0 (zero). Giving a negative integer as the tab index removes the component from the focus order entirely.

Managing Input Focus

81


User Interface Components

CSS Style Rules The component having the focus will have an additional style class with the -focus prefix. For example, a TextField would have style v-textfield-focus. For example (if we have the focusexample style defined for a parent of a text field), the following would make a text field blue when it has focus. .focusexample .v-textfield-focus { background: lightblue; }

5.4. Label Label is a text component that you can use to display non-editable text. The text will wrap around if the width of the containing component limits the length of the lines (except for preformatted text). // A container for the Label. Panel panel = new Panel("Panel Containing a Label"); panel.setWidth("200px"); // Defined width. main.addComponent(panel); panel.addComponent( new Label("This is a Label inside a Panel. There is enough " + "text in the label to make the text wrap if it " + "exceeds the width of the panel."));

As the size of the Panel in the above example is fixed, the text in the Label will wrap to fit the panel, as shown in Figure 5.16, â&#x20AC;&#x153;The Label Componentâ&#x20AC;?.

Figure 5.16. The Label Component

The contents of a label are formatted depending on the content mode. By default, the text is assumed to be plain text and any contained XML-specific characters will be quoted appropriately to allow rendering the contents of a label in XHTML in a web browser. The content mode can be set in the constructor or with setContentMode(), and can have the following values:

82

CONTENT_DEFAULT

The default content mode is CONTENT_TEXT (see below).

CONTENT_PREFORMATTED

Content mode, where the label contains preformatted text. It will be, by default, rendered with a fixed-width typewriter font. Preformatted text can contain line breaks, written in Java with the \n escape sequence for a newline character (ASCII 0x0a), or tabulator characters written with \t (ASCII 0x08).

CONTENT_RAW

Content mode, where the label contains raw text. Output is not required to be valid XML. It can be, for example, HTML, which can be unbalanced or otherwise invalid XML. The example below uses the <br> tag in HTML. While XHTML

Label


User Interface Components

should be preferred in most cases, this can be useful for some specific purposes where you may need to display loosely formatted HTML content. The raw mode also preserves character entities, some of which might otherwise be interpreted incorrectly. CONTENT_TEXT

Content mode, where the label contains only plain text. All characters are allowed, including the special <, >, and & characters in XML or HTML, which are quoted properly in XHTML while rendering the component. This is the default mode.

CONTENT_XHTML

Content mode, where the label contains XHTML.The content will be enclosed in a DIV element having the namespace "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd".

CONTENT_XML

Content mode, where the label contains well-formed and well-balanced XML. Each of the root elements must have their default namespace specified.

CONTENT_UIDL

Formatted content mode, where the contents are XML that is restricted to UIDL 1.0, the internal language of Vaadin for AJAX communications between the server and the browser. Obsolete since IT Mill Toolkit 5.0.

Warning Notice that the validity of XML or XHTML in a Label is not checked in the server when rendering the component and any errors can result in an error in the browser! You should validate the content before displaying it in the component, especially if it comes from an uncertain source. The following example demonstrates the use of Label in different modes. GridLayout labelgrid = new GridLayout (2,1); labelgrid.addComponent (new Label ("CONTENT_DEFAULT")); labelgrid.addComponent ( new Label ("This is a label in default mode: <plain text>", Label.CONTENT_DEFAULT)); labelgrid.addComponent (new Label ("CONTENT_PREFORMATTED")); labelgrid.addComponent ( new Label ("This is a preformatted label.\n"+ "The newline character \\n breaks the line.", Label.CONTENT_PREFORMATTED)); labelgrid.addComponent (new Label ("CONTENT_RAW")); labelgrid.addComponent ( new Label ("This is a label in raw mode.<br>It can contain, "+ "for example, unbalanced markup.", Label.CONTENT_RAW)); labelgrid.addComponent (new Label ("CONTENT_TEXT")); labelgrid.addComponent ( new Label ("This is a label in (plain) text mode", Label.CONTENT_TEXT)); labelgrid.addComponent (new Label ("CONTENT_XHTML")); labelgrid.addComponent (

Label

83


User Interface Components

new Label ("<i>This</i> is an <b>XHTML</b> formatted label", Label.CONTENT_XHTML)); labelgrid.addComponent (new Label ("CONTENT_XML")); labelgrid.addComponent ( new Label ("This is an <myelement>XML</myelement> "+ "formatted label", Label.CONTENT_XML)); main.addComponent(labelgrid);

The rendering will look as follows:

Figure 5.17. Label Modes Rendered on Screen

Using the XHTML, XML, or raw modes allow inclusion of, for example, images within the text flow, which is not possible with any regular layout components. The following example includes an image within the text flow, with the image coming from a class loader resource. ClassResource labelimage = new ClassResource ("labelimage.jpg", this); main.addComponent(new Label("Here we have an image <img src=\"" + this.getRelativeLocation(labelimage) + "\"/> within text.", Label.CONTENT_XHTML));

When you use a class loader resource, the image has to be included in the JAR of the web application. In this case, the labelimage.jpg needs to be in the default package. When rendered in a web browser, the output will look as follows:

Figure 5.18. Referencing An Image Resource in Label

Another solution would be to use the CustomLayout component, where you can write the component content as an XHTML fragment in a theme, but such a solution may be too heavy for most cases, and not flexible enough if the content needs to be dynamically generated. Notice that the rendering of XHTML depends on the assumption that the client software and the terminal adapter are XHTML based. It is possible to write a terminal adapter for a custom thin client application, which may not be able to render XHTML at all. There are also differences between web browsers in their support of XHTML.

84

Label


User Interface Components

Spacing with a Label You can use a Label to create vertical or horizontal space in a layout. If you need a empty "line" in a vertical layout, having just a label with empty text is not enough, as it will collapse to zero height. The same goes for a label with only whitespace as the label text. You need to use a nonbreaking space character, either &nbsp; or &#160;: layout.addComponent(new Label("&nbsp;", Label.CONTENT_XHTML));

Using the Label.CONTENT_PREFORMATTED mode has the same effect; preformatted spaces do not collapse in a vertical layout. In a HorizontalLayout, the width of a space character may be unpredictable if the label font is proportional, so you can use the preformatted mode to add em-width wide spaces. If you want a gap that has adjustable width or height, you can use an empty label if you specify a height or width for it. For example, to create vertical space in a VerticalLayout: Label gap = new Label(); gap.setHeight("1em"); verticalLayout.addComponent(gap);

You can make a flexible expanding spacer by having a relatively sized empty label with 100% height or width and setting the label as expanding in the layout. // A wide component bar HorizontalLayout horizontal = new HorizontalLayout(); horizontal.setWidth("100%"); // Have a component before the gap (a collapsing cell) Button button1 = new Button("I'm on the left"); horizontal.addComponent(button1); // An expanding gap spacer Label expandingGap = new Label(); expandingGap.setWidth("100%"); horizontal.addComponent(expandingGap); horizontal.setExpandRatio(expandingGap, 1.0f); // A component after the gap (a collapsing cell) Button button2 = new Button("I'm on the right"); horizontal.addComponent(button2);

CSS Style Rules The Label component has a v-label overall style. The Reindeer theme includes a number of predefined styles for typical formatting cases. These include "h1" (Reindeer.LABEL_H1) and "h2" (Reindeer.LABEL_H2) heading styles and "light" (Reindeer.LABEL_SMALL) style.

5.5. Link The Link component allows making references to resources that are either external or provided by the web server or by the application itself. While a Link appears like a hyperlink, it is not handled in the web browser. When a user clicks a link, the server receives an event and typically opens the referenced resource in the target window of the link. Resources are explained in Section 4.5, â&#x20AC;&#x153;Referencing Resourcesâ&#x20AC;?. Links to external resources can be made by using a URI as follows:

Spacing with a Label

85


User Interface Components

Link link = new Link ("link to a resource", new ExternalResource("http://www.vaadin.com/")));

With the simple contructor used in the above example, the link is opened in the current window. Using the constructor that takes the target window as a parameter, or by setting the window with setWindow, you can open the resource in another window, such as a native popup window or a FrameWindow. As the target window can be defined as a target string managed by the browser, the target can be any window, including windows not managed by the application itself. When the user clicks the link, the application will receive an event regarding the click and handle it to provide the resource. The link is therefore not an <a href> element in HTML and it does not have an URI. This has some additional consequences, such as that a link can not be marked as "visited" by the browser, unlike normal hyperlinks. If you wish to have an actual HTML anchor element, you need to customize the rendering of the component or use a Label with XHTML content mode and write the anchor element by yourself.

CSS Style Rules The Link component has v-link style by default. .v-link { }

When the mouse pointer hovers over the link, it will also have the over style.

5.6. TextField TextField is one of the most common user interface components and is highly versatile. It supports both single- and multi-line editing, password input, and buffering. The following example creates two simple text fields: a single-line and a multi-line TextField. /* Add a single-line text field. */ TextField subject = new TextField("Subject"); subject.setColumns(40); main.addComponent(subject); /* Add a multi-line text field. */ TextField message = new TextField("Message"); message.setRows(7); message.setColumns(40); main.addComponent(message);

Figure 5.19. Single- and Multi-Line Text Field Example

Notice how font size affects the width of the text fields even though the width was set with the same number of columns. This is a feature of HTML.

86

CSS Style Rules


User Interface Components

5.7. RichTextArea The RichTextArea field allows entering or editing formatted text. The toolbar provides all basic editing functionalities. The text content of RichTextArea is represented in HTML format. RichTextArea inherits TextField and does not add any API functionality over it. You can add new functionality by extending the client-side components VRichTextArea and VRichTextToolbar. As with TextField, the textual content of the rich text area is the Property of the field and can be set with setValue() and read with getValue(). // Create a rich text area final RichTextArea rtarea = new RichTextArea(); rtarea.setCaption("My Rich Text Area"); // Set initial content as HTML rtarea.setValue("<h1>Hello</h1>\n" + "<p>This rich text area contains some text.</p>");

Figure 5.20. Rich Text Area Component

Above, we used context-specific tags such as <h1> in the initial HTML content. The rich text area component does not allow creating such tags, only formatting tags, but it does preserve them unless the user edits them away. Any non-visible whitespace such as the new line character (\n) are removed from the content. For example, the value set above will be as follows when read from the field with getValue(): <h1>Hello</h1> <p>This rich text area contains some text.</p>

The rich text area is one of the few components in Vaadin that contain textual labels. The selection boxes in the toolbar are in English, and not be localized currently otherwise but by inheriting or reimplementing the client-side VRichTextToolbar widget. The buttons can be localized simply with CSS by downloading a copy of the toolbar background image, editing it, and replacing the default toolbar. The toolbar is a single image file from which the individual button icons are picked, so the order of the icons is different from the rendered. The image file depends on the client-side implementation of the toolbar. .v-richtextarea-richtextexample .gwt-ToggleButton .gwt-Image { background-image: url(img/richtextarea-toolbar-fi.png) !important; }

RichTextArea

87


User Interface Components

Figure 5.21. Regular English and a Localized Rich Text Area Toolbar

Cross-Site Scripting with RichTextArea The user input from a RichTextArea is transmitted as XHTML from the browser to server-side and is not sanitized. As the entire purpose of the RichTextArea component is to allow input of formatted text, you can not sanitize it just by removing all HTML tags. Also many attributes, such as style, should pass through the sanitization. See Section 11.10.1, “Sanitizing User Input to Prevent Cross-Site Scripting” for more details on Cross-Site scripting vulnerabilities and sanitization of user input.

CSS Style Rules .v-richtextarea { } .v-richtextarea .gwt-RichTextToolbar { } .v-richtextarea .gwt-RichTextArea { }

The rich text area consists of two main parts: the toolbar with overall style .gwt-RichTextToolbar and the editor area with style .gwt-RichTextArea.The editor area obviously contains all the elements and their styles that the HTML content contains. The toolbar contains buttons and drop-down list boxes with the following respective style names: .gwt-ToggleButton { } .gwt-ListBox { }

5.8. Date and Time Input with DateField The DateField component provides the means to display and input date and time. The field comes in two variations: PopupDateField, with a numeric input box and a popup calendar view, and InlineDateField, with the calendar view always visible. The DateField base class defaults to the popup variation. The example below illustrates the use of the DateField baseclass, which is equivalent to the PopupDateField. We set the initial time of the date field to current time by using the default constructor of the java.util.Date class. // Create a DateField with the default style DateField date = new DateField(); // Set the date and time to present date.setValue(new Date());

The result is shown in Figure 5.22, “DateField (PopupDateField) for Selecting Date and Time”.

88

Cross-Site Scripting with RichTextArea


User Interface Components

Figure 5.22. DateField (PopupDateField) for Selecting Date and Time

5.8.1. PopupDateField The PopupDateField provides date input using a text box for the date and time. As the DateField defaults to this component, the use is exactly the same as described earlier. Clicking the handle right of the date opens a popup view for selecting the year, month, and day, as well as time. Also the Down key opens the popup. Once opened, the user can navigate the calendar using the cursor keys. The date and time selected from the popup are displayed in the text box according to the default date and time format of the current locale, or as specified with setDateFormat(). The same format definitions are used for parsing user input.

Date and Time Format The date and time are normally displayed according to the default format for the current locale (see Section 5.3.5, “Locale”). You can specify a custom format with setDateFormat(). It takes a format string that follows the format of the SimpleDateFormat in Java. // Display only year, month, and day in ISO format date.setDateFormat("yyyy-MM-dd");

The result is shown in Figure 5.23, “Custom Date Format for PopupDateField”.

Figure 5.23. Custom Date Format for PopupDateField

The same format specification is also used for parsing user-input date and time, as described later.

Handling Malformed User Input A user can easily input a malformed or otherwise invalid date or time. DateField has two validation layers: first on the client-side and then on the server-side. The validity of the entered date is first validated on the client-side, immediately when the input box loses focus. If the date format is invalid, the v-datefield-parseerror style is set. Whether this causes a visible indication of a problem depends on the theme.The built-in reindeer

PopupDateField

89


User Interface Components

theme does not shown any indication by default, making server-side handling of the problem more convenient. .mydate.v-datefield-parseerror .v-textfield { background: pink; }

The setLenient(true) setting enables relaxed interpretation of dates, so that invalid dates, such as February 30th or March 0th, are wrapped to the next or previous month, for example. The server-side validation phase occurs when the date value is sent to the server. If the date field is set in immediate state, it occurs immediately after the field loses focus. Once this is done and if the status is still invalid, an error indicator is displayed beside the component. Hovering the mouse pointer over the indicator shows the error message. You can handle the errors by overriding the handleUnparsableDateString() method. The method gets the user input as a string parameter and can provide a custom parsing mechanism, as shown in the following example. // Create a date field with a custom parsing and a // custom error message for invalid format PopupDateField date = new PopupDateField("My Date") { @Override protected Date handleUnparsableDateString(String dateString) throws Property.ConversionException { // Try custom parsing String fields[] = dateString.split("/"); if (fields.length >= 3) { try { int year = Integer.parseInt(fields[0]); int month = Integer.parseInt(fields[1])-1; int day = Integer.parseInt(fields[2]); GregorianCalendar c = new GregorianCalendar(year, month, day); return c.getTime(); } catch (NumberFormatException e) { throw new Property. ConversionException("Not a number"); } } // Bad date throw new Property. ConversionException("Your date needs two slashes"); } }; // Display only year, month, and day in slash-delimited format date.setDateFormat("yyyy/MM/dd"); // Don't be too tight about the validity of dates // on the client-side date.setLenient(true);

The handler method must either return a parsed Date object or throw a ConversionException. Returning null will set the field value to null and clear the input box.

Customizing the Error Message In addition to customized parsing, overriding the handler method for unparseable input is useful for internationalization and other customization of the error message. You can also use it for another way for reporting the errors, as is done in the example below:

90

PopupDateField


User Interface Components

// Create a date field with a custom error message for invalid format PopupDateField date = new PopupDateField("My Date") { @Override protected Date handleUnparsableDateString(String dateString) throws Property.ConversionException { // Have a notification for the error getWindow().showNotification( "Your date needs two slashes", Notification.TYPE_WARNING_MESSAGE); // A failure must always also throw an exception throw new Property.ConversionException("Bad date"); } };

If the input is invalid, you should always throw the exception; returning a null value would make the input field empty, which is probably undesired.

Input Prompt Like other fields that have a text box, PopupDateField allows an input prompt that is visible until the user has input a value. You can set the prompt with setInputPrompt. PopupDateField date = new PopupDateField(); // Set the prompt date.setInputPrompt("Select a date"); // Set width explicitly to accommodate the prompt date.setWidth("10em");

The date field doesn't automatically scale to accommodate the prompt, so you need to set it explicitly with setWidth(). The input prompt is not available in the DateField superclass.

CSS Style Rules .v-datefield, v-datefield-popupcalendar {} .v-textfield, v-datefield-textfield {} .v-datefield-button {}

The top-level element of DateField and all its variants have v-datefield style. The base class and the PopupDateField also have the v-datefield-popupcalendar style. In addition, the top-level element has a style that indicates the resolution, with v-datefieldbasename and an extension, which is one of full, day, month, or year. The -full style is enabled when the resolution is smaller than a day. These styles are used mainly for controlling the appearance of the popup calendar. The text box has v-textfield and v-datefield-textfield styles, and the calendar button v-datefield-button. Once opened, the calendar popup has the following styles at the top level: .v-datefield-popup {} .v-popupcontent {} .v-datefield-calendarpanel {}

The top-level element of the floating popup calendar has .v-datefield-popup style. Observe that the popup frame is outside the HTML structure of the component, hence it is not enclosed

PopupDateField

91


User Interface Components

in the v-datefield element and does not include any custom styles. The content in the v-datefield-calendarpanel is the same as in InlineDateField, as described in Section 5.8.2, “InlineDateField”.

5.8.2. InlineDateField The InlineDateField provides a date picker component with a month view. The user can navigate months and years by clicking the appropriate arrows. Unlike with the popup variant, the month view is always visible in the inline field. // Create a DateField with the default style InlineDateField date = new InlineDateField(); // Set the date and time to present date.setValue(new java.util.Date());

The result is shown in Figure 5.24, “Example of the InlineDateField”.

Figure 5.24. Example of the InlineDateField

The user can also navigate the calendar using the cursor keys.

CSS Style Rules .v-datefield {} .v-datefield-calendarpanel {} .v-datefield-calendarpanel-header {} .v-datefield-calendarpanel-prevyear {} .v-datefield-calendarpanel-prevmonth {} .v-datefield-calendarpanel-month {} .v-datefield-calendarpanel-nextmonth {} .v-datefield-calendarpanel-nextyear {} .v-datefield-calendarpanel-body {} .v-datefield-calendarpanel-weekdays, .v-datefield-calendarpanel-weeknumbers {} .v-first {} .v-last {} .v-datefield-calendarpanel-weeknumber {} .v-datefield-calendarpanel-day {} .v-datefield-calendarpanel-time {} .v-datefield-time {} .v-select {} .v-label {}

The top-level element has the v-datefield style. In addition, the top-level element has a style name that indicates the resolution of the calendar, with v-datefield- basename and an extension, which is one of full, day, month, or year.The -full style is enabled when the resolution is smaller than a day.

92

InlineDateField


User Interface Components

The v-datefield-calendarpanel-weeknumbers and v-datefield-calendarpanel-weeknumber styles are enabled when the week numbers are enabled. The former controls the appearance of the weekday header and the latter the actual week numbers. The other style names should be self-explanatory. For weekdays, the v-first and v-last styles allow making rounded endings for the weekday bar.

5.8.3. Time Resolution You probably will not need milliseconds in most applications, and might not even need the time, but just the date, or month or year. The visibility of the input components is controlled by time resolution, which can be set with setResolution() method.The method takes as its parameters the lowest visible component, typically RESOLUTION_DAY for just dates and RESOLUTION_MIN for dates with time in hours and minutes. Please see the API Reference for a complete list of resolution parameters.

5.8.4. DateField Locale The date and time are displayed according to the locale of the user, as reported by the browser. You can set a custom locale with the setLocale() method of AbstractComponent, as described in Section 5.3.5, â&#x20AC;&#x153;Localeâ&#x20AC;?. Only Gregorian calendar is supported.

5.9. Button The Button is a user interface component that is normally used for finalizing input and initiating some action. When the user clicks a button, a Button.ClickEvent is emitted. A listener that inherits the Button.ClickListener interface can handle clicks with the buttonClick() method. public class TheButton extends CustomComponent implements Button.ClickListener { Button thebutton; public TheButton() { // Create a Button with the given caption. thebutton = new Button ("Do not push this button"); // Listen for ClickEvents. thebutton.addListener(this); setCompositionRoot(thebutton); } /** Handle click events for the button. */ public void buttonClick (Button.ClickEvent event) { thebutton.setCaption ("Do not push this button again"); } }

Figure 5.25. An Example of a Button

As a user interface often has several buttons, you can differentiate between them either by comparing the Button object reference returned by the getButton() method of Button.ClickEvent to a kept reference or by using a separate listener method for each button. The listening

Time Resolution

93


User Interface Components

object and method can be given to the constructor. For a detailed description of these patterns together with some examples, please see Section 3.5, “Events and Listeners”.

CSS Style Rules .v-button { }

The exact CSS style name can be different if a Button has the switchMode attribute enabled. See the alternative CSS styles below. Adding the "small" style name enables a smaller style for the Button. You can also use the BUTTON_SMALL constant in Runo and Reindeer theme classes as well. The BaseTheme class also has a BUTTON_LINK style, with "link" style name, which makes the button look like a hyperlink.

5.10. CheckBox CheckBox is a two-state selection component that can be either checked or unchecked. The caption of the check box will be placed right of the actual check box. Vaadin provides two ways to create check boxes: individual check boxes with the CheckBox component described in this section and check box groups with the OptionGroup component in multiple selection mode, as described in Section 5.11.5, “Radio Button and Check Box Groups with OptionGroup”. Clicking on a check box will change its state. The state is the Boolean property of the Button, and can be set with setValue() and obtained with getValue() method of the Property interface. Changing the value of a check box will cause a ValueChangeEvent, which can be handled by a ValueChangeListener. // A check box with default state (not checked, false). final CheckBox checkbox1 = new CheckBox("My CheckBox"); main.addComponent(checkbox1); // Another check box with explicitly set checked state. final CheckBox checkbox2 = new CheckBox("Checked CheckBox"); checkbox2.setValue(true); main.addComponent(checkbox2); // Make some application logic. We use anonymous listener // classes here. The above references were defined as final // to allow accessing them from inside anonymous classes. checkbox1.addListener(new ValueChangeListener() { public void valueChange(ValueChangeEvent event) { // Copy the value to the other checkbox. checkbox2.setValue(checkbox1.getValue()); } }); checkbox2.addListener(new ValueChangeListener() { public void valueChange(ValueChangeEvent event) { // Copy the value to the other checkbox. checkbox1.setValue(checkbox2.getValue()); } });

Figure 5.26. An Example of a Check Box

94

CSS Style Rules


User Interface Components

For an example on the use of check boxes in a table, see Section 5.12, â&#x20AC;&#x153;Tableâ&#x20AC;?.

CSS Style Rules .v-checkbox { }

5.11. Selecting Items Vaadin gives many alternatives for selecting one or more items from a list, using drop-down and regular lists, radio button and check box groups, tables, trees, and so on. The core library includes the following selection components, all based on the AbstractSelect class: Select

In single selection mode, a drop-down list with a text input area, which the user can use to filter the displayed items. In multiselect mode, a list box equivalent to ListSelect.

ComboBox

A drop-down list for single selection. Otherwise as Select, but the user can also enter new items. The component also provides an input prompt.

ListSelect

A vertical list box for selecting items in either single or multiple selection mode.

NativeSelect

Provides selection using the native selection component of the browser, typically a drop-down list for single selection and a multi-line list in multiselect mode. This uses the <select> element in HTML.

OptionGroup

Shows the items as a vertically arranged group of radio buttons in the single selection mode and of check boxes in multiple selection mode.

TwinColSelect

Shows two list boxes side by side where the user can select items from a list of available items and move them to a list of selected items using control buttons.

In addition, the Tree and Table components allow special forms of selection. They also inherit the AbstractSelect.

5.11.1. Binding Selection Components to Data The selection components are strongly coupled with the Vaadin Data Model.The selectable items in all selection components are objects that implement the Item interface and are contained in a Container. The current selection is bound to the Property interface. Even though the data model is used, the selection components allow simple use in the most common cases. Each selection component is bound to a default container type, which supports management of items without need to implement a container. See Chapter 9, Binding Components to Data for a detailed description of the data model, its interfaces, and built-in implementations.

Adding New Items New items are added with the addItem() method defined in the Container interface.

CSS Style Rules

95


User Interface Components

// Create a selection component Select select = new Select ("Select something here"); // Add some items and give each an item ID select.addItem("Mercury"); select.addItem("Venus"); select.addItem("Earth");

The addItem() method creates an empty Item, which is identified by its item identifier (IID) object, given as the parameter. This item ID is by default used also as the caption of the item, as explained in the next section. The identifier is typically a String. The item is of a type specific to the container and has itself little relevance for most selection components, as the properties of an item may not be used in any way (except in Table), only the item ID. The item identifier can be of any object type. We could as well have given integers for the item identifiers and set the captions explicitly with setItemCaption(). You could also add an item with the parameterless addItem(), which returns an automatically generated item ID. // Create a selection component Select select = new Select("My Select"); // Add an item with a generated ID Object itemId = select.addItem(); select.setItemCaption(itemId, "The Sun"); // Select the item select.setValue(itemId);

Some container types may support passing the actual data object to the add method. For example, you can add items to a BeanItemContainer with addBean(). Such implementations can use a separate item ID object, or the data object itself as the item ID, as is done in addBean(). In the latter case you can not depend on the default way of acquiring the item caption; see the description of the different caption modes later. The following section describes the different options for determining the item captions.

Item Captions The displayed captions of items in a selection component can be set explicitly with setItemCaption() or determined from the item IDs or item properties. This behaviour is defined with the caption mode, which you can set with setItemCaptionMode(). The default mode is ITEM_CAPTION_MODE_EXPLICIT_DEFAULTS_ID, which uses the item identifiers for the captions, unless given explicitly. In addition to a caption, an item can have an icon. The icon is set with setItemIcon().

Caption Modes for Selection Components ITEM_CAPTION_MODE_EXPLICIT_DEFAULTS_ID This is the default caption mode and its flexibility allows using it in most cases. By default, the item identifier will be used as the caption. The identifier object does not necessarily have to be a string; the caption is retrieved with toString() method. If the caption is specified explicitly with setItemCaption(), it overrides the item identifier. Select select = new Select("Moons of Mars"); // Use the item ID also as the caption of this item

96

Binding Selection Components to Data


User Interface Components

select.addItem(new Integer(1)); // Set item caption for this item explicitly select.addItem(2); // same as "new Integer(2)" select.setItemCaption(2, "Deimos");

ITEM_CAPTION_MODE_EXPLICIT Captions must be explicitly specified with setItemCaption(). If they are not, the caption will be empty. Such items with empty captions will nevertheless be displayed in the Select component as empty items. If they have an icon, they will be visible. ITEM_CAPTION_MODE_ICON_ONLY Only icons are shown, captions are hidden. ITEM_CAPTION_MODE_ID

String representation of the item identifier object is used as caption. This is useful when the identifier is a string, and also when the identifier is an complex object that has a string representation. For example: Select select = new Select("Inner Planets"); select.setItemCaptionMode(Select.ITEM_CAPTION_MODE_ID);

// A class that implements toString() class PlanetId extends Object implements Serializable { String planetName; PlanetId (String name) { planetName = name; } public String toString () { return "The Planet " + planetName; } } // Use such objects as item identifiers String planets[] = {"Mercury", "Venus", "Earth", "Mars"}; for (int i=0; i<planets.length; i++) select.addItem(new PlanetId(planets[i]));

ITEM_CAPTION_MODE_INDEX

Index number of item is used as caption. This caption mode is applicable only to data sources that implement the Container.Indexed interface. If the interface is not available, the component will throw a ClassCastException. The Select component itself does not implement this interface, so the mode is not usable without a separate data source. An IndexedContainer, for example, would work.

ITEM_CAPTION_MODE_ITEM

String representation of item, acquired with toString(), is used as the caption.This is applicable mainly when using a custom Item class, which also requires using a custom Container that is used as a data source for the Select component.

ITEM_CAPTION_MODE_PROPERTY Item captions are read from the String representation of the property with the identifier specified with setItemCaptionPropertyId(). This is useful, for

Binding Selection Components to Data

97


User Interface Components

example, when you have a container that you use as the data source for a Select, and you want to use a specific property for caption. In the example below, we bind a selection component to a bean container and use a property of the bean as the caption. /* A bean with a "name" property. */ public class Planet implements Serializable { String name; public Planet(String name) { this.name = name; } public void setName(String name) { this.name = name; } public String getName() { return name; } } void propertyModeExample() { VerticalLayout layout = new VerticalLayout(); // Have a bean container to put the beans in BeanItemContainer<Planet> container = new BeanItemContainer<Planet>(Planet.class); // Put some example data in it container.addItem(new Planet("Mercury")); container.addItem(new Planet("Venus")); container.addItem(new Planet("Earth")); container.addItem(new Planet("Mars")); // Create a selection component bound to the container Select select = new Select("Planets", container); // Set the caption mode to read the caption directly // from the 'name' property of the bean select.setItemCaptionMode( Select.ITEM_CAPTION_MODE_PROPERTY); select.setItemCaptionPropertyId("name"); layout.addComponent(select);

Getting and Setting Selection A selection component provides the current selection as the property of the component (with the Property interface). The property value is an item identifier object that identifies the selected item. You can get the identifier with getValue() of the Property interface. You can select an item with the corresponding setValue() method. In multiselect mode, the property will be an unmodifiable set of item identifiers. If no item is selected, the property will be null in single selection mode or an empty collection in multiselect mode.

98

Binding Selection Components to Data


User Interface Components

The Select and NativeSelect components will show "-" selection when no actual item is selected. This is the null selection item identifier. You can set an alternative ID with setNullSelectionItemId(). Setting the alternative null ID is merely a visual text; the getValue() will still return null value if no item is selected, or an empty set in multiselect mode. The item identifier of the currently selected item will be set as the property of the Select object. You can access it with the getValue() method of the Property interface of the component. Also, when handling changes in a Select component with the Property.ValueChangeListener interface, the Property.ValueChangeEvent will have the selected item as the property of the event, accessible with the getProperty() method.

Figure 5.27. Selected Item

5.11.2. Basic Select Component The Select component allows, in single selection mode, selecting an item from a drop-down list. The component also has a text field area, which allows entering search text by which the items shown in the drop-down list are filtered. In multiple selection mode, the component shows the items in a vertical list box, identical to ListSelect.

Figure 5.28. The Select Component

Filtered Selection The Select component allows filtering the items available for selection. The component shows as an input box for entering text. The text entered in the input box is used for filtering the available items shown in a drop-down list. Pressing Enter will complete the item in the input box. Pressing Up- and Down-arrows can be used for selecting an item from the drop-down list. The drop-down

Basic Select Component

99


User Interface Components

list is paged and clicking on the scroll buttons will change to the next or previous page. The list selection can also be done with the arrow keys on the keyboard. The shown items are loaded from the server as needed, so the number of items held in the component can be quite large. Vaadin provides two filtering modes: FILTERINGMODE_CONTAINS matches any item that contains the string given in the text field part of the component and FILTERINGMODE_STARTSWITH matches only items that begin with the given string. The filtering mode is set with setFilteringMode(). Setting the filtering mode to the default value FILTERINGMODE_OFF disables filtering. Select select = new Select("Enter containing substring"); select.setFilteringMode(AbstractSelect.Filtering.FILTERINGMODE_CONTAINS); /* Fill the component with some items. */ final String[] planets = new String[] { "Mercury", "Venus", "Earth", "Mars", "Jupiter", "Saturn", "Uranus", "Neptune" }; for (int i = 0; i < planets.length; i++) for (int j = 0; j < planets.length; j++) { select.addItem(planets[j] + " to " + planets[i]);

The above example uses the containment filter that matches to all items containing the input string. As shown in Figure 5.29, â&#x20AC;&#x153;Filtered Selectionâ&#x20AC;? below, when we type some text in the input area, the drop-down list will show all the matching items.

Figure 5.29. Filtered Selection

CSS Style Rules .v-filterselect { } .v-filterselect-input { } .v-filterselect-button { } .v-filterselect-suggestpopup { } .v-filterselect-prefpage-off { } .v-filterselect-suggestmenu { } .v-filterselect-status { } .v-select { } .v-select-select { }

100

Basic Select Component


User Interface Components

In its default state, only the input field of the Select component is visible. The entire component is enclosed in v-filterselect style, the input field has v-filterselect-input style and the button in the right end that opens and closes the drop-down result list has v-filterselect-button style. The drop-down result list has an overall v-filterselect-suggestpopup style. It contains the list of suggestions with v-filterselect-suggestmenu style and a status bar in the bottom with v-filterselect-status style. The list of suggestions is padded with an area with v-filterselect-prefpage-off style above and below the list. In multiselect-mode, the styles of the component aere identical to ListSelect component, with v-select overall style and v-select-select for the native selection element.

5.11.3. ListSelect The ListSelect component is list box that shows the selectable items in a vertical list. If the number of items exceeds the height of the component, a scrollbar is shown.The component allows both single and multiple selection modes, which you can set with setMultiSelect(). It is visually identical in both modes. // Create the selection component ListSelect select = new ListSelect("My Selection"); // Add some items select.addItem("Mercury"); select.addItem("Venus"); select.addItem("Earth"); ... select.setNullSelectionAllowed(false); // Show 5 items and a scrollbar if there are more select.setRows(5);

The number of visible items is set with setRows().

Figure 5.30. The ListSelect Component

CSS Style Rules .v-select {} .v-select-select {}

The component has a v-select overall style. The native select element has v-select-select style.

ListSelect

101


User Interface Components

5.11.4. Native Selection Component NativeSelect NativeSelect offers the native selection component of web browsers, using the HTML <select> element. The component is shown as a drop-down list. // Create the selection component final NativeSelect select = new NativeSelect("Native Selection"); // Add some items select.addItem("Mercury"); select.addItem("Venus"); ... // Set the width in "columns" as in TextField select.setColumns(10); select.setNullSelectionAllowed(false);

The setColumns() allows setting the width of the list as "columns", which is a measure that depends on the browser.

Figure 5.31. The NativeSelect Component

Multiple selection mode is not allowed; you should use the ListSelect component instead. Also adding new items, which would be enabled with setNewItemsAllowed(), is not allowed.

CSS Style Rules .v-select {} .v-select-select {}

The component has a v-select overall style. The native select element has v-select-select style.

5.11.5. Radio Button and Check Box Groups with OptionGroup The OptionGroup class provides selection from alternatives using a group of radio buttons in single selection mode. In multiple selection mode, the items show up as check boxes. OptionGroup optiongroup = new OptionGroup("My Option Group"); // Use the multiple selection mode. myselect.setMultiSelect(true);

102

Native Selection Component NativeSelect


User Interface Components

Figure 5.32, “Option Button Group in Single and Multiple Selection Mode” shows the OptionGroup in both single and multiple selection mode.

Figure 5.32. Option Button Group in Single and Multiple Selection Mode

You can create check boxes individually using the CheckBox class, as described in Section 5.10, “CheckBox”. The advantages of the OptionGroup component are that as it maintains the individual check box objects, you can get an array of the currently selected items easily, and that you can easily change the appearance of a single component.

Disabling Items You can disable individual items in an OptionGroup with setItemEnabled(). The user can not select or deselect disabled items in multi-select mode, but in single-select mode the use can change the selection from a disabled to an enabled item. The selections can be changed programmatically regardless of whether an item is enabled or disabled. You can find out whether an item is enabled with isItemEnabled(). The setItemEnabled() identifies the item to be disabled by its item ID. // Have an option group OptionGroup group = new OptionGroup("My Disabled Group"); group.addItem("One"); group.addItem("Two"); group.addItem("Three"); // Disable one item group.setItemEnabled("Two", false);

The item IDs are also used for the captions in this example. The result is shown in Figure 5.33, “OptionGroup with a Disabled Item”.

Figure 5.33. OptionGroup with a Disabled Item

Setting an item as disabled turns on the v-disabled style for it.

CSS Style Rules .v-select-optiongroup {} .v-select-option.v-checkbox {} .v-select-option.v-radiobutton {}

Radio Button and Check Box Groups with OptionGroup

103


User Interface Components

The v-select-optiongroup is the overall style for the component. Each check box will have the v-checkbox style, borrowed from the CheckBox component, and each radio button the v-radiobutton style. Both the radio buttons and check boxes will also have the v-select-option style that allows styling regardless of the option type. Disabled items have additionally the v-disabled style. The options are normally laid out vertically. You can use horizontal layout by setting display: inline-block for the options. Notice, however, that the options will wrap when they reach the available width in the containing layout, and the layout must not have an undefined width. .v-select-option { display: inline-block; }

5.11.6. Twin Column Selection with TwinColSelect The TwinColSelect class provides a multiple selection component that shows two lists side by side. The user can select items from the list on the left and click on the ">>" button to move them to the list on the right. Items can be moved back by selecting them and clicking on the "<<" button.

Figure 5.34. Twin Column Selection

CSS Style Rules .v-select-twincol {} .v-select-twincol-options {} .v-select-twincol-selections {} .v-select-twincol-buttons {} .v-select-twincol-deco {}

5.11.7. Allowing Adding New Items The selection components allow the user to add new items, with a user interface similar to combo boxes in desktop user interfaces. You need to enable the newItemsAllowed mode with the setNewItemsAllowed() method. myselect.setNewItemsAllowed(true);

The user interface for adding new items depends on the selection component and the selection mode. The regular Select component in single selection mode, which appears as a combo box, allows you to simply type the new item in the combo box and hit Enter to add it. In most other selection components, as well as in the multiple selection mode of the regular Select component, a text field that allows entering new items is shown below the selection list, and clicking the + button will add the item in the list, as illustrated in Figure 5.35, â&#x20AC;&#x153;Select Component with Adding New Items Allowedâ&#x20AC;?.

104

Twin Column Selection with TwinColSelect


User Interface Components

Figure 5.35. Select Component with Adding New Items Allowed

The identifier of an item added by the user will be a String object identical to the caption of the item. You should consider this if the item identifier of automatically filled items is some other type or otherwise not identical to the caption. Adding new items is possible in both single and multiple selection modes and in all styles. Adding new items may not be possible if the Select is bound to an external Container that does not allow adding new items.

5.11.8. Multiple Selection Mode Setting the Select, NativeSelect, or OptionGroup components to multiple selection mode with the setMultiSelect() method changes their appearance to allow selecting multiple items. Select and NativeSelect

These components appear as a native HTML selection list, as shown in Figure 5.35, “Select Component with Adding New Items Allowed”. By holding the Ctrl or Shift key pressed, the user can select multiple items.

OptionGroup

The option group, which is a radio button group in single selection mode, will show as a check box group in multiple selection mode. See Section 5.11.5, “Radio Button and Check Box Groups with OptionGroup”.

The TwinColSelect, described in Section 5.11.6, “Twin Column Selection with TwinColSelect”, is a special multiple selection mode that is not meaningful for single selection. myselect.setMultiSelect(true);

As in single selection mode, the selected items are set as the property of the Select object. In multiple selection mode, the property is a Collection of currently selected items. You can get and set the property with the getValue() and setValue() methods as usual. A change in the selection will trigger a ValueChangeEvent, which you can handle with a Propery.ValueChangeListener. As usual, you should use setImmediate(true) to trigger the event immediately when the user changes the selection. The following example shows how to handle selection changes with a listener. public class SelectExample extends CustomComponent implements Property.ValueChangeListener { // Create a Select object with a caption. Select select = new Select("This is a Select component");

Multiple Selection Mode

105


User Interface Components

VerticalLayout layout = new VerticalLayout(); Label status = new Label("-"); SelectExample () { setCompositionRoot (layout); layout.addComponent(select); // Fill the component with some items. final String[] planets = new String[] { "Mercury", "Venus", "Earth", "Mars", "Jupiter", "Saturn", "Uranus", "Neptune"}; for (int i=0; i<planets.length; i++) select.addItem(planets[i]); // By default, the change event is not triggered // immediately when the selection changes. // This enables the immediate events. select.setImmediate(true); // Listen for changes in the selection. select.addListener(this); layout.addComponent(status); } /* Respond to change in the selection. */ public void valueChange(Property.ValueChangeEvent event) { // The event.getProperty() returns the Item ID (IID) // of the currently selected item in the component. status.setValue("Currently selected item ID: " + event.getProperty()); } }

5.11.9. Other Common Features Item Icons You can set an icon for each item with setItemIcon(), or define an item property that provides the icon resource with setItemIconPropertyId(), in a fashion similar to captions. Notice, however, that icons are not supported in NativeSelect, TwinColSelect, and some other selection components and modes.This is because HTML does not support images inside the native select elements. Icons are also not really visually applicable.

5.12. Table The Table component is intended for presenting tabular data organized in rows and columns. The Table is one of the most versatile components in Vaadin. Table cells can include text or arbitrary UI components. You can easily implement editing of the table data, for example clicking on a cell could change it to a text field for editing. The data contained in a Table is managed using the Data Model of Vaadin (see Chapter 9, Binding Components to Data), through the Container interface of the Table.This makes it possible to bind a table directly to a data source, such as a database query. Only the visible part of the table is loaded into the browser and moving the visible window with the scrollbar loads content from the server. While the data is being loaded, a tooltip will be displayed that shows the current range and total number of items in the table. The rows of the table are items in the container and the columns are properties. Each table row (item) is identified with an item identifier (IID), and each column (property) with a property identifier (PID).

106

Other Common Features


User Interface Components

When creating a table, you first need to define columns with addContainerProperty(). This method comes in two flavors. The simpler one takes the property ID of the column and uses it also as the caption of the column. The more complex one allows differing PID and header for the column. This may make, for example, internationalization of table headers easier, because if a PID is internationalized, the internationalization has to be used everywhere where the PID is used. The complex form of the method also allows defining an icon for the column from a resource. The "default value" parameter is used when new properties (columns) are added to the table, to fill in the missing values. (This default has no meaning in the usual case, such as below, where we add items after defining the properties.) /* Create the table with a caption. */ Table table = new Table("This is my Table"); /* Define the names and data types of columns. * The "default value" parameter is meaningless here. */ table.addContainerProperty("First Name", String.class, null); table.addContainerProperty("Last Name", String.class, null); table.addContainerProperty("Year", Integer.class, null); /* Add a few items in the table. */ table.addItem(new Object[] { "Nicolaus","Copernicus",new Integer(1473)}, table.addItem(new Object[] { "Tycho", "Brahe", new Integer(1546)}, table.addItem(new Object[] { "Giordano","Bruno", new Integer(1548)}, table.addItem(new Object[] { "Galileo", "Galilei", new Integer(1564)}, table.addItem(new Object[] { "Johannes","Kepler", new Integer(1571)}, table.addItem(new Object[] { "Isaac", "Newton", new Integer(1643)},

new Integer(1)); new Integer(2)); new Integer(3)); new Integer(4)); new Integer(5)); new Integer(6));

In this example, we used an increasing Integer object as the Item Identifier, given as the second parameter to addItem(). The actual rows are given simply as object arrays, in the same order in which the properties were added. The objects must be of the correct class, as defined in the addContainerProperty() calls.

Figure 5.36. Basic Table Example

Scalability of the Table is largely dictated by the container. The default IndexedContainer is relatively heavy and can cause scalability problems, for example, when updating the values. Use of an optimized application-specific container is recommended. Table does not have a limit for the number of items and is just as fast with hundreds of thousands of items as with just a few. With the current implementation of scrolling, there is a limit of around 500 000 rows, depending on the browser and the pixel height of rows.

Table

107


User Interface Components

5.12.1. Selecting Items in a Table The Table allows selecting one or more items by clicking them with the mouse. When the user selects an item, the IID of the item will be set as the property of the table and a ValueChangeEvent is triggered. To enable selection, you need to set the table selectable. You will also need to set it as immediate in most cases, as we do below, because without it, the change in the property will not be communicated immediately to the server. The following example shows how to enable the selection of items in a Table and how to handle ValueChangeEvent events that are caused by changes in selection. You need to handle the event with the valueChange() method of the Property.ValueChangeListener interface. // Allow selecting items from the table. table.setSelectable(true); // Send changes in selection immediately to server. table.setImmediate(true); // Shows feedback from selection. final Label current = new Label("Selected: -"); // Handle selection change. table.addListener(new Property.ValueChangeListener() { public void valueChange(ValueChangeEvent event) { current.setValue("Selected: " + table.getValue()); } });

Figure 5.37. Table Selection Example

If the user clicks on an already selected item, the selection will deselected and the table property will have null value. You can disable this behaviour by setting setNullSelectionAllowed(false) for the table. The selection is the value of the table's property, so you can get it with getValue(). You can get it also from a reference to the table itself. In single selection mode, the value is the item identifier of the selected item or null if no item is selected. In multiple selection mode (see below), the value is a Set of item identifiers. Notice that the set is unmodifiable, so you can not simply change it to change the selection.

Multiple Selection Mode A table can also be in multiselect mode, where a user can select multiple items by clicking them with left mouse button while holding the Ctrl key (or Meta key) pressed. If Ctrl is not held, clicking an item will select it and other selected items are deselected. The user can select a range by selecting an item, holding the Shift key pressed, and clicking another item, in which case all the items between the two are also selected. Multiple ranges can be selected by first selecting a

108

Selecting Items in a Table


User Interface Components

range, then selecting an item while holding Ctrl, and then selecting another item with both Ctrl and Shift pressed. The multiselect mode is enabled with the setMultiSelect() method of the Select interface of Table. Setting table in multiselect mode does not implicitly set it as selectable, so it must be set separately. The setMultiSelectMode() property affects the control of multiple selection: MultiSelectMode.DEFAULT is the default behaviour, which requires holding the Ctrl (or Meta) key pressed while selecting items, while in MultiSelectMode.SIMPLE holding the Ctrl key is not needed. In the simple mode, items can only be deselected by clicking them.

5.12.2. Table Features Page Length and Scrollbar The default style for Table provides a table with a scrollbar. The scrollbar is located at the right side of the table and becomes visible when the number of items in the table exceeds the page length, that is, the number of visible items.You can set the page length with setPageLength(). Setting the page length to zero makes all the rows in a table visible, no matter how many rows there are. Notice that this also effectively disables buffering, as all the entire table is loaded to the browser at once. Using such tables to generate reports does not scale up very well, as there is some inevitable overhead in rendering a table with Ajax. For very large reports, generating HTML directly is a more scalable solution.

Resizing Columns You can set the width of a column programmatically from the server-side with setColumnWidth(). The column is identified by the property ID and the width is given in pixels. The user can resize table columns by dragging the resize handle between two columns. Resizing a table column causes a ColumnResizeEvent, which you can handle with a Table.ColumnResizeListener. The table must be set in immediate mode if you want to receive the resize events immediately, which is typical. table.addListener(new Table.ColumnResizeListener() { public void columnResize(ColumnResizeEvent event) { // Get the new width of the resized column int width = event.getCurrentWidth(); // Get the property ID of the resized column String column = (String) event.getPropertyId(); // Do something with the information table.setColumnFooter(column, String.valueOf(width) + "px"); } }); // Must be immediate to send the resize events immediately table.setImmediate(true);

See Figure 5.38, â&#x20AC;&#x153;Resizing Columnsâ&#x20AC;? for a result after the columns of a table has been resized.

Table Features

109


User Interface Components

Figure 5.38. Resizing Columns

Reordering Columns If setColumnReorderingAllowed(true) is set, the user can reorder table columns by dragging them with the mouse from the column header,

Collapsing Columns When setColumnCollapsingAllowed(true) is set, the right side of the table header shows a drop-down list that allows selecting which columns are shown. Collapsing columns is different than hiding columns with setVisibleColumns(), which hides the columns completely so that they can not be made visible (uncollapsed) from the user interface. You can collapse columns programmatically with setColumnCollapsed(). Collapsing must be enabled before collapsing columns with the method or it will throw an IllegalAccessException. // Allow the user to collapse and uncollapse columns table.setColumnCollapsingAllowed(true); // Collapse this column programmatically try { table.setColumnCollapsed("born", true); } catch (IllegalAccessException e) { // Can't occur - collapsing was allowed above System.err.println("Something horrible occurred"); } // Give enough width for the table to accommodate the // initially collapsed column later table.setWidth("250px");

See Figure 5.39, â&#x20AC;&#x153;Collapsing Columnsâ&#x20AC;?.

Figure 5.39. Collapsing Columns

If the table has undefined width, it minimizes its width to fit the width of the visible columns. If some columns are initially collapsed, the width of the table may not be enough to accomodate them later, which will result in an ugly horizontal scrollbar. You should consider giving the table enough width to accomodate columns uncollapsed by the user.

110

Table Features


User Interface Components

Components Inside a Table The cells of a Table can contain any user interface components, not just strings. If the rows are higher than the row height defined in the default theme, you have to define the proper row height in a custom theme. When handling events for components inside a Table, such as for the Button in the example below, you usually need to know the item the component belongs to. Components do not themselves know about the table or the specific item in which a component is contained. Therefore, the handling method must use some other means for finding out the Item ID of the item. There are a few possibilities. Usually the easiest way is to use the setData() method to attach an arbitrary object to a component.You can subclass the component and include the identity information there.You can also simply search the entire table for the item with the component, although that solution may not be so scalable. The example below includes table rows with a Label in XHTML formatting mode, a multiline TextField, a CheckBox, and a Button that shows as a link. // Create a table and add a style to allow setting the row height in theme. final Table table = new Table(); table.addStyleName("components-inside"); /* Define the names and data types of columns. * The "default value" parameter is meaningless here. */ table.addContainerProperty("Sum", Label.class, table.addContainerProperty("Is Transferred", CheckBox.class, table.addContainerProperty("Comments", TextField.class, table.addContainerProperty("Details", Button.class,

null); null); null); null);

/* Add a few items in the table. */ for (int i=0; i<100; i++) { // Create the fields for the current table row Label sumField = new Label(String.format( "Sum is <b>$%04.2f</b><br/><i>(VAT incl.)</i>", new Object[] {new Double(Math.random()*1000)}), Label.CONTENT_XHTML); CheckBox transferredField = new CheckBox("is transferred"); // Multiline text field. This required modifying the // height of the table row. TextField commentsField = new TextField(); commentsField.setRows(3); // The Table item identifier for the row. Integer itemId = new Integer(i); // Create a button and handle its click. A Button does not // know the item it is contained in, so we have to store the // item ID as user-defined data. Button detailsField = new Button("show details"); detailsField.setData(itemId); detailsField.addListener(new Button.ClickListener() { public void buttonClick(ClickEvent event) { // Get the item identifier from the user-defined data. Integer itemId = (Integer)event.getButton().getData(); getWindow().showNotification("Link "+ itemId.intValue()+" clicked."); } }); detailsField.addStyleName("link"); // Create the table row. table.addItem(new Object[] {sumField, transferredField,

Table Features

111


User Interface Components

commentsField, detailsField}, itemId); } // Show just three rows because they are so high. table.setPageLength(3);

The row height has to be set higher than the default with a style rule such as the following: /* Table rows contain three-row TextField components. */ .v-table-components-inside .v-table-cell-content { height: 54px; }

The table will look as shown in Figure 5.40, “Components in a Table”.

Figure 5.40. Components in a Table

Iterating Over a Table As the items in a Table are not indexed, iterating over the items has to be done using an iterator. The getItemIds() method of the Container interface of Table returns a Collection of item identifiers over which you can iterate using an Iterator. For an example about iterating over a Table, please see Section 9.4, “Collecting items in Containers”. Notice that you may not modify the Table during iteration, that is, add or remove items. Changing the data is allowed.

5.12.3. Editing the Values in a Table Normally, a Table simply displays the items and their fields as text. If you want to allow the user to edit the values, you can either put them inside components as we did above, or you can simply call setEditable(true) and the cells are automatically turned into editable fields. Let us begin with a regular table with a some columns with usual Java types, namely a Date, Boolean, and a String. // Create a table. It is by default not editable. final Table table = new Table(); // Define the names and data types of columns. table.addContainerProperty("Date", Date.class, null); table.addContainerProperty("Work", Boolean.class, null); table.addContainerProperty("Comments", String.class, null); // Add a few items in the table. for (int i=0; i<100; i++) { Calendar calendar = new GregorianCalendar(2008,0,1);

112

Editing the Values in a Table


User Interface Components

calendar.add(Calendar.DAY_OF_YEAR, i); // Create the table row. table.addItem(new Object[] {calendar.getTime(), new Boolean(false), ""}, new Integer(i)); // Item identifier } table.setPageLength(8); layout.addComponent(table);

You could put the table in editable mode right away if you need to. We'll continue the example by adding a mechanism to switch the Table from and to the editable mode. final CheckBox switchEditable = new CheckBox("Editable"); switchEditable.addListener(new Property.ValueChangeListener() { public void valueChange(ValueChangeEvent event) { table.setEditable(((Boolean)event.getProperty() .getValue()).booleanValue()); } }); switchEditable.setImmediate(true); layout.addComponent(switchEditable);

Now, when you check to checkbox, the components in the table turn into editable fields, as shown in Figure 5.41, â&#x20AC;&#x153;A Table in Normal and Editable Modeâ&#x20AC;?.

Figure 5.41. A Table in Normal and Editable Mode

Field Factories The field components that allow editing the values of particular types in a table are defined in a field factory that implements the TableFieldFactory interface. The default implementation is DefaultFieldFactory, which offers the following crude mappings:

Table 5.2. Type to Field Mappings in DefaultFieldFactory Property Type Mapped to Field Class Date

A DateField.

Boolean

A CheckBox.

Item

A Form. The fields of the form are automatically created from the item's properties using a FormFieldFactory. The normal use for this property type is inside a Form and is less useful inside a Table.

other

A TextField.The text field manages conversions from the basic types, if possible.

Editing the Values in a Table

113


User Interface Components

Field factories are covered with more detail in Section 5.17.2, â&#x20AC;&#x153;Binding Form to Dataâ&#x20AC;?. You could just implement the TableFieldFactory interface, but we recommend that you extend the DefaultFieldFactory according to your needs. In the default implementation, the mappings are defined in the createFieldByPropertyType() method (you might want to look at the source code) both for tables and forms.

Navigation in Editable Mode In the editable mode, the editor fields can have focus. Pressing Tab moves the focus to next column or, at the last column, to the first column of the next item. Respectively, pressing Shift-Tab moves the focus backward. If the focus is in the last column of the last visible item, the pressing Tab moves the focus outside the table. Moving backward from the first column of the first item moves the focus to the table itself. Some updates to the table, such as changing the headers or footers or regenerating a column, can move the focus from an editor component to the table itself. The default behaviour may be undesirable in many cases. For example, the focus also goes through any read-only editor fields and can move out of the table inappropriately.You can provide better navigation is to use event handler for shortcut keys such as Tab, Arrow Up, Arrow Down, and Enter. // Keyboard navigation class KbdHandler implements Handler { Action tab_next = new ShortcutAction("Tab", ShortcutAction.KeyCode.TAB, null); Action tab_prev = new ShortcutAction("Shift+Tab", ShortcutAction.KeyCode.TAB, new int[] {ShortcutAction.ModifierKey.SHIFT}); Action cur_down = new ShortcutAction("Down", ShortcutAction.KeyCode.ARROW_DOWN, null); Action cur_up = new ShortcutAction("Up", ShortcutAction.KeyCode.ARROW_UP, null); Action enter = new ShortcutAction("Enter", ShortcutAction.KeyCode.ENTER, null); public Action[] getActions(Object target, Object sender) { return new Action[] {tab_next, tab_prev, cur_down, cur_up, enter}; } public void handleAction(Action action, Object sender, Object target) { if (target instanceof TextField) { // Move according to keypress int itemid = (Integer) ((TextField) target).getData(); if (action == tab_next || action == cur_down) itemid++; else if (action == tab_prev || action == cur_up) itemid--; // On enter, just stay where you were. If we did // not catch the enter action, the focus would be // moved to wrong place. if (itemid >= 0 && itemid < table.size()) { TextField newTF = valueFields.get(itemid); if (newTF != null) newTF.focus(); } } } } // Panel that handles keyboard navigation Panel navigator = new Panel(); navigator.addStyleName(Reindeer.PANEL_LIGHT);

114

Editing the Values in a Table


User Interface Components

navigator.addComponent(table); navigator.addActionHandler(new KbdHandler());

The main issue in implementing keyboard navigation in an editable table is that the editor fields do not know the table they are in. To find the parent table, you can either look up in the component container hierarchy or simply store a reference to the table with setData() in the field component. The other issue is that you can not acquire a reference to an editor field from the Table component. One solution is to use some external collection, such as a HashMap, to map item IDs to the editor fields. // Can't access the editable components from the table so // must store the information final HashMap<Integer,TextField> valueFields = new HashMap<Integer,TextField>();

The map has to be filled in a TableFieldFactory, such as in the following. You also need to set the reference to the table there and you can also set the initial focus there. table.setTableFieldFactory(new TableFieldFactory () { public Field createField(Container container, Object itemId, Object propertyId, Component uiContext) { TextField field = new TextField((String) propertyId); // User can only edit the numeric column if (propertyId.equals("Source of Fear")) field.setReadOnly(true); else { // The numeric column // The field needs to know the item it is in field.setData(itemId); // Remember the field valueFields.put((Integer) itemId, field); // Focus the first editable value if (((Integer)itemId) == 0) field.focus(); } return field; } });

The issues are complicated by the fact that the editor fields are not generated for the entire table, but only for a cache window that includes the visible items and some items above and below it. For example, if the beginning of a big scrollable table is visible, the editor component for the last item does not exist. This issue is relevant mostly if you want to have wrap-around navigation that jumps from the last to first item and vice versa.

5.12.4. Column Headers and Footers Table supports both column headers and footers; the headers are enabled by default.

Headers The table header displays the column headers at the top of the table. You can use the column headers to reorder or resize the columns, as described earlier. By default, the header of a column is the property ID of the column, unless given explicitly with setColumnHeader(). // Define the properties table.addContainerProperty("lastname", String.class, null); table.addContainerProperty("born", Integer.class, null); table.addContainerProperty("died", Integer.class, null);

Column Headers and Footers

115


User Interface Components

// Set nicer header names table.setColumnHeader("lastname", "Name"); table.setColumnHeader("born", "Born"); table.setColumnHeader("died", "Died");

The text of the column headers and the visibility of the header depends on the column header mode. The header is visible by default, but you can disable it with setColumnHeaderMode(Table.COLUMN_HEADER_MODE_HIDDEN).

Footers The table footer can be useful for displaying sums or averages of values in a column, and so on. The footer is not visible by default; you can enable it with setFooterVisible(true). Unlike in the header, the column headers are empty by default. You can set their value with setColumnFooter(). The columns are identified by their property ID. The following example shows how to calculate average of the values in a column: // Have a table with a numeric column Table table = new Table("Custom Table Footer"); table.addContainerProperty("Name", String.class, null); table.addContainerProperty("Died At Age", Integer.class, null); // Insert some data Object people[][] = {{"Galileo", 77}, {"Monnier", 83}, {"Vaisala", 79}, {"Oterma", 86}}; for (int i=0; i<people.length; i++) table.addItem(people[i], new Integer(i)); // Calculate the average of the numeric column double avgAge = 0; for (int i=0; i<people.length; i++) avgAge += (Integer) people[i][1]; avgAge /= people.length; // Set the footers table.setFooterVisible(true); table.setColumnFooter("Name", "Average"); table.setColumnFooter("Died At Age", String.valueOf(avgAge)); // Adjust the table height a bit table.setPageLength(table.size());

The resulting table is shown in Figure 5.42, â&#x20AC;&#x153;A Table with a Footerâ&#x20AC;?.

Figure 5.42. A Table with a Footer

116

Column Headers and Footers


User Interface Components

Handling Mouse Clicks on Headers and Footers Normally, when the user clicks a column header, the table will be sorted by the column, assuming that the data source is Sortable and sorting is not disabled. In some cases, you might want some other functionality when the user clicks the column header, such as selecting the column in some way. Clicks in the header cause a HeaderClickEvent, which you can handle with a Table.HeaderClickListener. Click events on the table header (and footer) are, like button clicks, sent immediately to server, so there is no need to set setImmediate(). // Handle the header clicks table.addListener(new Table.HeaderClickListener() { public void headerClick(HeaderClickEvent event) { String column = (String) event.getPropertyId(); getWindow().showNotification("Clicked " + column + "with " + event.getButtonName()); } }); // Disable the default sorting behavior table.setSortDisabled(true);

Setting a click handler does not automatically disable the sorting behavior of the header; you need to disable it explicitly with setSortDisabled(true). Header click events are not sent when the user clicks the column resize handlers to drag them. The HeaderClickEvent object provides the identity of the clicked column with getPropertyId(). The getButton() reports the mouse button with which the click was made: BUTTON_LEFT, BUTTON_RIGHT, or BUTTON_MIDDLE. The getButtonName() a human-readable button name in English: "left", "right", or "middle". The isShiftKey(), isCtrlKey(), etc., methods indicate if the Shift, Ctrl, Alt or other modifier keys were pressed during the click. Clicks in the footer cause a FooterClickEvent, which you can handle with a Table.FooterClickListener. Footers do not have any default click behavior, like the sorting in the header. Otherwise, handling clicks in the footer is equivalent to handling clicks in the header.

5.12.5. Generated Table Columns You might want to have a column that has values calculated from other columns. Or you might want to format table columns in some way, for example if you have columns that display currencies. The ColumnGenerator interface allows defining custom generators for such columns. You add new generated columns to a Table with addGeneratedColumn(). It takes the column identifier as its parameters. Usually you want to have a more user-friendly and possibly internationalized column header. You can set the header and a possible icon by calling addContainerProperty() before adding the generated column. // Define table columns. table.addContainerProperty( "date", Date.class, table.addContainerProperty( "quantity", Double.class, table.addContainerProperty( "price", Double.class, table.addContainerProperty( "total", Double.class,

null, "Date",

null, null);

null, "Quantity (l)", null, null); null, "Price (e/l)",

null, null);

null, "Total (e)",

null, null);

// Define the generated columns and their generators.

Generated Table Columns

117


User Interface Components

table.addGeneratedColumn("date", new DateColumnGenerator()); table.addGeneratedColumn("quantity", new ValueColumnGenerator("%.2f l")); table.addGeneratedColumn("price", new PriceColumnGenerator()); table.addGeneratedColumn("total", new ValueColumnGenerator("%.2f e"));

Notice that the addGeneratedColumn() always places the generated columns as the last column, even if you defined some other order previously. You will have to set the proper order with setVisibleColumns(). table.setVisibleColumns(new Object[] {"date", "quantity", "price", "total"});

The generators are objects that implement the Table.ColumnGenerator interface and its generateCell() method.The method gets the identity of the item and column as its parameters, in addition to the table object. It has to return a component object. The following example defines a generator for formatting Double valued fields according to a format string (as in java.util.Formatter). /** Formats the value in a column containing Double objects. */ class ValueColumnGenerator implements Table.ColumnGenerator { String format; /* Format string for the Double values. */ /** * Creates double value column formatter with the given * format string. */ public ValueColumnGenerator(String format) { this.format = format; } /** * Generates the cell containing the Double value. * The column is irrelevant in this use case. */ public Component generateCell(Table source, Object itemId, Object columnId) { // Get the object stored in the cell as a property Property prop = source.getItem(itemId).getItemProperty(columnId); if (prop.getType().equals(Double.class)) { Label label = new Label(String.format(format, new Object[] { (Double) prop.getValue() })); // Set styles for the column: one indicating that it's // a value and a more specific one with the column // name in it. This assumes that the column name // is proper for CSS. label.addStyleName("column-type-value"); label.addStyleName("column-" + (String) columnId); return label; } return null; } }

The generator is called for all the visible (or more accurately cached) items in a table. If the user scrolls the table to another position in the table, the columns of the new visible rows are generated dynamically. The columns in the visible (cached) rows are also generated always when an item has a value change. It is therefore usually safe to calculate the value of generated cells from the values of different rows (items).

118

Generated Table Columns


User Interface Components

When you set a table as editable, regular fields will change to editing fields. When the user changes the values in the fields, the generated columns will be updated automatically. Putting a table with generated columns in editable mode has a few quirks. The editable mode of Table does not affect generated columns. You have two alternatives: either you generate the editing fields in the generator or, in case of formatter generators, remove the generator in the editable mode. The example below uses the latter approach. // Have a check box that allows the user // to make the quantity and total columns editable. final CheckBox editable = new CheckBox( "Edit the input values - calculated columns are regenerated"); editable.setImmediate(true); editable.addListener(new ClickListener() { public void buttonClick(ClickEvent event) { table.setEditable(editable.booleanValue()); // The columns may not be generated when we want to // have them editable. if (editable.booleanValue()) { table.removeGeneratedColumn("quantity"); table.removeGeneratedColumn("total"); } else { // Not editable // Show the formatted values. table.addGeneratedColumn("quantity", new ValueColumnGenerator("%.2f l")); table.addGeneratedColumn("total", new ValueColumnGenerator("%.2f e")); } // The visible columns are affected by removal // and addition of generated columns so we have // to redefine them. table.setVisibleColumns(new Object[] {"date", "quantity", "price", "total", "consumption", "dailycost"}); } });

You will also have to set the editing fields in immediate mode to have the update occur immediately when an edit field loses the focus. You can set the fields in immediate mode with the a custom TableFieldFactory, such as the one given below, that just extends the default implementation to set the mode: public class ImmediateFieldFactory extends DefaultFieldFactory { public Field createField(Container container, Object itemId, Object propertyId, Component uiContext) { // Let the DefaultFieldFactory create the fields... Field field = super.createField(container, itemId, propertyId, uiContext); // ...and just set them as immediate. ((AbstractField)field).setImmediate(true); return field; } } ... table.setFieldFactory(new ImmediateFieldFactory());

If you generate the editing fields with the column generator, you avoid having to use such a field factory, but of course have to generate the fields for both normal and editable modes.

Generated Table Columns

119


User Interface Components

Figure 5.43, “Table with Generated Columns in Normal and Editable Mode” shows a table with columns calculated (blue) and simply formatted (black) with column generators.

Figure 5.43. Table with Generated Columns in Normal and Editable Mode

You can find the complete generated columns example in the Feature Browser demo application in the installation package, in com.vaadin.demo.featurebrowser.GeneratedColumnExample.java.

5.12.6. Formatting Table Columns The displayed values of properties shown in a table are normally formatted using the toString() method of each property. Customizing the format of a column can be done in several ways: • Using ColumnGenerator to generate a second column that is formatted. The original column needs to be set invisible. See Section 5.12.5, “Generated Table Columns”. • Using a PropertyFormatter as a proxy between the table and the data property. This also normally requires using an mediate container in the table. • Overriding the default formatPropertyValue() in Table. As using a PropertyFormatter is generally much more awkward than overriding the formatPropertyValue(), its use is not described here. You can override formatPropertyValue() as is done in the following example: // Create a table that overrides the default // property (column) format final Table table = new Table("Formatted Table") { @Override protected String formatPropertyValue(Object rowId,

120

Formatting Table Columns


User Interface Components

Object colId, Property property) { // Format by property type if (property.getType() == Date.class) { SimpleDateFormat df = new SimpleDateFormat("yyyy-MM-dd hh:mm:ss"); return df.format((Date)property.getValue()); } return super.formatPropertyValue(rowId, colId, property); } }; // The table has some columns table.addContainerProperty("Time", Date.class, null); ... Fill the table with data ...

You can also distinguish between columns by the colId parameter, which is the property ID of the column. DecimalFormat is useful for formatting decimal values. ... in formatPropertyValue() ... } else if (pid.equals("Value")) { // Format a decimal value for a specific locale DecimalFormat df = new DecimalFormat("#.00", new DecimalFormatSymbols(locale)); return df.format((Double) property.getValue()); } ... table.addContainerProperty("Value", Double.class, null);

A table with the formatted date and decimal value columns is shown in Figure 5.44, “Formatted Table Columns”.

Figure 5.44. Formatted Table Columns

You can use CSS for further styling of table rows, columns, and individual cells by using a CellStyleGenerator. It is described in Section 5.12.7, “CSS Style Rules”.

5.12.7. CSS Style Rules Styling the overall style of a Table can be done with the following CSS rules. .v-table {} .v-table-header-wrap {} .v-table-header {} .v-table-header-cell {} .v-table-resizer {} /* Column resizer handle. */ .v-table-caption-container {} .v-table-body {} .v-table-row-spacer {} .v-table-table {}

CSS Style Rules

121


User Interface Components

.v-table-row {} .v-table-cell-content {}

Notice that some of the widths and heights in a table are calculated dynamically and can not be set in CSS.

Setting Individual Cell Styles The Table.CellStyleGenerator interface allows you to set the CSS style for each individual cell in a table. You need to implement the getStyle(), which gets the row (item) and column (property) identifiers as parameters and can return a style name for the cell. The returned style name will be concatenated to prefix "v-table-cell-content-". Alternatively, you can use a Table.ColumnGenerator (see Section 5.12.5, â&#x20AC;&#x153;Generated Table Columnsâ&#x20AC;?) to generate the actual UI components of the cells and add style names to them. A cell style generator is not used for the cells in generated columns. Table table = new Table("Table with Cell Styles"); table.addStyleName("checkerboard"); // Add some columns in the table. In this example, the property // IDs of the container are integers so we can determine the // column number easily. table.addContainerProperty("0", String.class, null, "", null, null); for (int i=0; i<8; i++) table.addContainerProperty(""+(i+1), String.class, null, String.valueOf((char) (65+i)), null, null); // Add some items in the table. table.addItem(new Object[]{ "1", "R", "N", "B", "Q", "K", "B", "N", "R"}, new table.addItem(new Object[]{ "2", "P", "P", "P", "P", "P", "P", "P", "P"}, new for (int i=2; i<6; i++) table.addItem(new Object[]{String.valueOf(i+1), "", "", "", "", "", "", "", ""}, new table.addItem(new Object[]{ "7", "P", "P", "P", "P", "P", "P", "P", "P"}, new table.addItem(new Object[]{ "8", "R", "N", "B", "Q", "K", "B", "N", "R"}, new table.setPageLength(8);

Integer(0)); Integer(1));

Integer(i)); Integer(6)); Integer(7));

// Set cell style generator table.setCellStyleGenerator(new Table.CellStyleGenerator() { public String getStyle(Object itemId, Object propertyId) { int row = ((Integer)itemId).intValue(); int col = Integer.parseInt((String)propertyId); // The first column. if (col == 0) return "rowheader"; // Other cells. if ((row+col)%2 == 0) return "black"; else return "white"; } });

You can then style the cells, for example, as follows: /* Center the text in header. */ .v-table-header-cell {

122

CSS Style Rules


User Interface Components

text-align: center; } /* Basic style for all cells. */ .v-table-checkerboard .v-table-cell-content { text-align: center; vertical-align: middle; padding-top: 12px; width: 20px; height: 28px; } /* Style specifically for the row header cells. */ .v-table-cell-content-rowheader { background: #E7EDF3 url(../default/table/img/header-bg.png) repeat-x scroll 0 0; } /* Style specifically for the "white" cells. */ .v-table-cell-content-white { background: white; color: black; } /* Style specifically for the "black" cells. */ .v-table-cell-content-black { background: black; color: white; }

The table will look as shown in Figure 5.45, “Cell Style Generator for a Table”.

Figure 5.45. Cell Style Generator for a Table

5.13. Tree The Tree component allows a natural way to represent data that has hierarchical relationships, such as filesystems or message threads. The Tree component in Vaadin works much like the tree components of most modern desktop user interface toolkits, for example in directory browsing. The typical use of the Tree component is for displaying a hierachical menu, like a menu on the left side of the screen, as in Figure 5.46, “A Tree Component as a Menu”, or for displaying

Tree

123


User Interface Components

filesystems or other hierarchical datasets. The menu style makes the appearance of the tree more suitable for this purpose. final Object[][] planets = new Object[][]{ new Object[]{"Mercury"}, new Object[]{"Venus"}, new Object[]{"Earth", "The Moon"}, new Object[]{"Mars", "Phobos", "Deimos"}, new Object[]{"Jupiter", "Io", "Europa", "Ganymedes", "Callisto"}, new Object[]{"Saturn", "Titan", "Tethys", "Dione", "Rhea", "Iapetus"}, new Object[]{"Uranus", "Miranda", "Ariel", "Umbriel", "Titania", "Oberon"}, new Object[]{"Neptune", "Triton", "Proteus", "Nereid", "Larissa"}}; Tree tree = new Tree("The Planets and Major Moons"); /* Add planets as root items in the tree. */ for (int i=0; i<planets.length; i++) { String planet = (String) (planets[i][0]); tree.addItem(planet); if (planets[i].length == 1) { // The planet has no moons so make it a leaf. tree.setChildrenAllowed(planet, false); } else { // Add children (moons) under the planets. for (int j=1; j<planets[i].length; j++) { String moon = (String) planets[i][j]; // Add the item as a regular item. tree.addItem(moon); // Set it to be a child. tree.setParent(moon, planet); // Make the moons look like leaves. tree.setChildrenAllowed(moon, false); } // Expand the subtree. tree.expandItemsRecursively(planet); } } main.addComponent(tree);

Figure 5.46, â&#x20AC;&#x153;A Tree Component as a Menuâ&#x20AC;? below shows the tree from the code example in a practical situation. You can read or set the currently selected item by the value property of the Tree component, that is, with getValue() and setValue(). When the user clicks an item on a tree, the tree will receive an ValueChangeEvent, which you can catch with a ValueChangeListener. To receive the event immediately after the click, you need to set the tree as setImmediate(true). The Tree component uses Container data sources much like the Table component, with the addition that it also utilizes hierarchy information maintained by a HierarchicalContainer. The contained items can be of any item type supported by the container. The default container and its addItem() assume that the items are strings and the string value is used as the item ID.

124

Tree


User Interface Components

Figure 5.46. A Tree Component as a Menu

5.14. MenuBar The MenuBar component allows creating horizontal dropdown menus, much like the main menu in desktop applications. // Create a menu bar final MenuBar menubar = new MenuBar(); main.addComponent(menubar);

You insert the top-level menu items to a MenuBar object with the addItem() method. It takes a string label, an icon resource, and a command as its parameters. The icon and command are not required and can be null. MenuBar.MenuItem beverages = menubar.addItem("Beverages", null, null);

The command is called when the user clicks the item. A menu command is a class that implements the MenuBar.Command interface. // A feedback component final Label selection = new Label("-"); main.addComponent(selection); // Define a common menu command for all the menu items. MenuBar.Command mycommand = new MenuBar.Command() { public void menuSelected(MenuItem selectedItem) { selection.setValue("Ordered a " + selectedItem.getText() + " from menu."); } };

The addItem() method returns a MenuBar.MenuItem object, which you can use to add submenu items. The MenuItem has an identical addItem() method. // Put some items in the menu hierarchically MenuBar.MenuItem beverages = menubar.addItem("Beverages", null, null); MenuBar.MenuItem hot_beverages = beverages.addItem("Hot", null, null); hot_beverages.addItem("Tea", null, mycommand); hot_beverages.addItem("Coffee", null, mycommand); MenuBar.MenuItem cold_beverages = beverages.addItem("Cold", null, null);

MenuBar

125


User Interface Components

cold_beverages.addItem("Milk", null, mycommand); // Another top-level item MenuBar.MenuItem snacks = menubar.addItem("Snacks", null, null); snacks.addItem("Weisswurst", null, mycommand); snacks.addItem("Salami", null, mycommand); // Yet another top-level item MenuBar.MenuItem services = menubar.addItem("Services", null, null); services.addItem("Car Service", null, mycommand);

The menu will look as follows:

Figure 5.47. Menu Bar

CSS Style Rules .v-menubar { } .gwt-MenuItem {} .gwt-MenuItem-selected {}

The menu bar has the overall style name .v-menubar. Each menu item has .gwt-MenuItem style normally and .gwt-MenuItem-selected when the item is selected.

5.15. Embedded The Embedded component allows displaying embedded media objects, such as images, animations, or any embeddable media type supported by the browser. The contents of an Embedded component are managed as resources. For documentation on resources, see Section 4.5, â&#x20AC;&#x153;Referencing Resourcesâ&#x20AC;?. The following example displays an image from the same Java package as the class itself using the class loader. Embedded image = new Embedded("Yes, logo:", new ClassResource("vaadin-logo.png", this)); main.addComponent(image);

126

CSS Style Rules


User Interface Components

Figure 5.48. Embedded Image

The Embedded component supports several different content types, which are rendered differently in HTML. You can set the content type with setType(), although for images, as in the above example, the type is determined automatically. Embedded.TYPE_OBJECT

The default embedded type, allows embedding certain file types inside HTML <object> and <embed> elements.

Embedded.TYPE_IMAGE

Embeds an image inside a HTML <img> element.

Embedded.TYPE_BROWSER

Embeds a browser frame inside a HTML <iframe> element.

5.15.1. Embedded Objects The Embedded.TYPE_OBJECT is the default and most generic embedded type, which allows embedding media objects inside HTML <object> and <embed> elements. You need define the MIME type for the object type. Currently, only Shockwave Flash application/x-shockwave-flash).

animations

are

supported

(MIME

type

// Create a Shockware Flash resource final ClassResource flashResource = new ClassResource("itmill_spin.swf", getApplication()); // Display the resource in a Embedded compoant final Embedded embedded = new Embedded("Embedded Caption", flashResource); // This is the default type, but we set it anyway. embedded.setType(Embedded.TYPE_OBJECT); // This is recorgnized automatically, but set it anyway. embedded.setMimeType("application/x-shockwave-flash");

You can set object parameters with setParameter(), which takes a parameter's name and value as strings. The object parameters are included in the HTML as <param> elements.

5.15.2. Embedded Images Images are embedded with the type Embedded.TYPE_IMAGE, although you do not normally need to set the type explicitly, as it is recognized automatically from the MIME type of the resource, as in the example above.

Embedded Objects

127


User Interface Components

You can find another example of displaying an image from FileResource in Section 5.16, “Upload”. Another example, in Section 4.5.5, “Stream Resources”, shows how you can generate the content of an Embedded component dynamically using a StreamResource. If you have a dynamically generated image, for example with a StreamResource, and the data changes, you need to reload the image in the browser. Because of how caching is handled in some browsers, you are best off by renaming the filename of the resource with a unique name, such as one including a timestamp. You should set cache time to zero with setCacheTime() for the resource object when you create it. // Create the stream resource with some initial filename. StreamResource imageResource = new StreamResource(imageSource, "initial-filename.png", getApplication()); // Instruct browser not to cache the image. imageResource.setCacheTime(0); // Display the image in an Embedded component. Embedded embedded = new Embedded("", imageResource);

When refreshing, you also need to call requestRepaint() for the Embedded object. // This needs to be done, but is not sufficient. embedded.requestRepaint(); // Generate a filename with a timestamp. SimpleDateFormat df = new SimpleDateFormat("yyyyMMddHHmmssSSS"); String filename = "myfilename-" + df.format(new Date()) + ".png"; // Replace the filename in the resource. imageResource.setFilename(makeImageFilename());

You can find more detailed information about the StreamResource in Section 4.5.5, “Stream Resources”.

5.15.3. Browser Frames The browser frame type allows you to embed external content inside an HTML <iframe> element. You can refer to a URL with an ExternalResource object. URLs are given with the standard Java URL class. URL url = new URL("http://dev.vaadin.com/"); Embedded browser = new Embedded("", new ExternalResource(url)); browser.setType(Embedded.TYPE_BROWSER); main.addComponent(browser);

5.16. Upload The Upload component allows a user to upload files to the server. It displays a file name entry box, a file selection button, and an upload submit button. The user can either write the filename in the text area or click the Browse button to select a file. After the file is selected, the user sends the file by pressing the upload submit button. // Create the Upload component. Upload upload = new Upload("Upload the file here", this);

128

Browser Frames


User Interface Components

Figure 5.49. Upload Component

You can set the text of the upload button with setButtonCaption(), as in the example above, but it is difficult to change the look of the Browse button. This is a security feature of web browsers. The language of the Browse button is determined by the browser, so if you wish to have the language of the Upload component consistent, you will have to use the same language in your application. upload.setButtonCaption("Upload Now");

The uploaded files are typically stored as files in a file system, in a database, or as temporary objects in memory. The upload component writes the received data to an java.io.OutputStream so you have plenty of freedom in how you can process the upload content. To use the Upload component, you need to define a class that implements the Upload.Receiver interface. The receiveUpload() method is called when the user clicks the submit button. The method must return an OutputStream. To do this, it typically creates a File or a memory buffer where the stream is written. The method gets the file name and MIME type of the file, as reported by the browser. When an upload is finished, successfully or unsuccessfully, the Upload component will emit the Upload.FinishedEvent event.To receive it, you need to implement the Upload.FinishedListener interface, and register the listening object in the Upload component. The event object will also include the file name, MIME type, and length of the file. Notice that the more specific Upload.FailedEvent and Upload.SucceededEvent events will be called in the cases where the upload failed or succeeded, respectively. The following example allows uploading images to /tmp/uploads directory in (UNIX) filesystem (the directory must exist or the upload fails). The component displays the last uploaded image in an Embedded component. import import import import import

java.io.File; java.io.FileOutputStream; java.io.OutputStream; com.vaadin.terminal.FileResource; com.vaadin.ui.*;

public class MyUploader extends CustomComponent implements Upload.SucceededListener, Upload.FailedListener, Upload.Receiver { Panel root; Panel imagePanel; File file;

// Root element for contained components. // Panel that contains the uploaded image. // File to write to.

MyUploader() { root = new Panel("My Upload Component"); setCompositionRoot(root); // Create the Upload component. final Upload upload = new Upload("Upload the file here", this); // Use a custom button caption instead of plain "Upload". upload.setButtonCaption("Upload Now"); // Listen for events regarding the success of upload.

Upload

129


User Interface Components

upload.addListener((Upload.SucceededListener) this); upload.addListener((Upload.FailedListener) this); root.addComponent(upload); root.addComponent(new Label("Click 'Browse' to "+ "select a file and then click 'Upload'.")); // Create a panel for displaying the uploaded image. imagePanel = new Panel("Uploaded image"); imagePanel.addComponent( new Label("No image uploaded yet")); root.addComponent(imagePanel); } // Callback method to begin receiving the upload. public OutputStream receiveUpload(String filename, String MIMEType) { FileOutputStream fos = null; // Output stream to write to file = new File("/tmp/uploads/" + filename); try { // Open the file for writing. fos = new FileOutputStream(file); } catch (final java.io.FileNotFoundException e) { // Error while opening the file. Not reported here. e.printStackTrace(); return null; } return fos; // Return the output stream to write to } // This is called if the upload is finished. public void uploadSucceeded(Upload.SucceededEvent event) { // Log the upload on screen. root.addComponent(new Label("File " + event.getFilename() + " of type '" + event.getMIMEType() + "' uploaded.")); // Display the uploaded file in the image panel. final FileResource imageResource = new FileResource(file, getApplication()); imagePanel.removeAllComponents(); imagePanel.addComponent(new Embedded("", imageResource)); } // This is called if the upload fails. public void uploadFailed(Upload.FailedEvent event) { // Log the failure on screen. root.addComponent(new Label("Uploading " + event.getFilename() + " of type '" + event.getMIMEType() + "' failed.")); } }

The example does not check the type of the uploaded files in any way, which will cause an error if the content is anything else but an image. The program also assumes that the MIME type of the file is resolved correctly based on the file name extension. After uploading an image, the component will look as show in Figure 5.50, â&#x20AC;&#x153;Image Upload Exampleâ&#x20AC;? below. The browser shows the Browse button localized.

130

Upload


User Interface Components

Figure 5.50. Image Upload Example

5.17. Form Most web applications need forms. The Form component in Vaadin offers an easy way to create forms where the fields can be automatically generated from a data source that is bound to the form. The BeanItem adapter allows the data sources to be just JavaBeans or Plain Old Java Objects (POJOs) with just the setter and getter methods. Form manages buffering so that the form contents can be committed to the data source only when filling the form is complete, and before that, the user can discard any changes. The Form component is also a layout, with a bounding box, a caption, a description field, and a special error indicator. As such, it can also be used within logical forms to group input fields.

5.17.1. Form as a User Interface Component To begin with the Form, it is a UI component with a layout suitable for its purpose. A Form has a caption, a description, a layout that contains the fields, an error indicator, and a footer, as illustrated in Figure 5.51, â&#x20AC;&#x153;Layout of the Form Componentâ&#x20AC;? below. Unlike with other components, the caption is shown within the border. (See the details below on how to enable the border with CSS, as it may not be enabled in the default style.)

Figure 5.51. Layout of the Form Component

Form

131


User Interface Components

Unlike most components, Form does not accept the caption in the constructor, as forms are often captionless, but you can give the caption with the setCaption(). While the description text, which you can set with setDescription(), is shown as a tooltip in most other components, a Form displays it in top of the form box as shown in the figure above. Form form = new Form(); form.setCaption("Form Caption"); form.setDescription("This is a description of the Form that is " + "displayed in the upper part of the form. You normally " + "enter some descriptive text about the form and its " + "use here.");

Form has FormLayout as its default layout, but you can set any other layout with setContent(). See Section 6.5, “FormLayout” for more information. Note that the Form itself handles layout for the description, the footer and other common elements of the form. The user-set layout only manages the contained fields and their captions. The Form is most of all a container for fields so it offers many kinds of automation for creating and managing fields. You can, of course, create fields directly in the layout, but it is usually more desirable to bind the fields to the connected data source. // Add a field directly to the layout. This field will // not be bound to the data source Item of the form. form.getLayout().addComponent(new TextField("A Field")); // Add a field and bind it to an named item property. form.addField("another", new TextField("Another Field"));

Binding forms and their fields to data objects is described further in Section 5.17.2, “Binding Form to Data” below. The Form has a special error indicator inside the form. The indicator can show the following types of error messages: • Errors set with the setComponentError() method of the form. For example: form.setComponentError(new UserError("This is the error indicator of the Form."));

• Errors caused by a validator attached to the Form with addValidator(). • Errors caused by validators attached to the fields inside forms, if setValidationVisible(true) is set for the form. This type of validation is explained futher in Section 5.17.3, “Validating Form Input” below. • Errors from automatic validation of fields set as required with setRequired(true) if an error message has also been set with setRequiredError(). Only a single error is displayed in the error indicator at a time. Finally, Form has a footer area. The footer is a HorizontalLayout by default, but you can change it with setFooter(). // Set the footer layout. form.setFooter(new VerticalLayout()); form.getFooter().addComponent( new Label("This is the footer area of the Form. "+ "You can use any layout here. "+ "This is nice for buttons."));

132

Form as a User Interface Component


User Interface Components

// Have a button bar in the footer. HorizontalLayout okbar = new HorizontalLayout(); okbar.setHeight("25px"); form.getFooter().addComponent(okbar); // Add an Ok (commit), Reset (discard), and Cancel buttons // for the form. Button okbutton = new Button("OK", form, "commit"); okbar.addComponent(okbutton); okbar.setComponentAlignment(okbutton, Alignment.TOP_RIGHT); okbar.addComponent(new Button("Reset", form, "discard")); okbar.addComponent(new Button("Cancel"));

CSS Style Rules .v-form {} .v-form legend .v-form fieldset {} .v-form-error {} .v-form-errormessage {} .v-form-description {}

The top-level style name of a Form component is v-form. It is important to notice that the form is implemented as a HTML <fieldset>, which allows placing the caption (or "legend") inside the border. It would not be so meaningful to set a border for the top-level form element. The following example sets a border around the form, as is done in Figure 5.51, â&#x20AC;&#x153;Layout of the Form Componentâ&#x20AC;? above. .v-form fieldset { border: thin solid; }

The top-level element of the form has the style name v-form-error if a component error has been set for the form.

5.17.2. Binding Form to Data The main purpose of the Form component is that you can bind it to a data source and let the Form generate and manage fields automatically. The data source can be any class that implements the Item interface, which is part of the Vaadin Data Model, as described in Chapter 9, Binding Components to Data. You can either implement the Item interface yourself, which can be overly complicated, or use the ready BeanItem adapter to bind the form to any JavaBean object. You can also use PropertysetItem to bind the form to an ad hoc set of Property objects, resembling a Map. Let us consider the following simple JavaBean with proper setter and getter methods for the member variables. /** A simple JavaBean. */ public class PersonBean { String name; String city; public void setName(String name) { this.name = name; } public String getName() { return name; } public void setCity(String city) {

Binding Form to Data

133


User Interface Components

this.city = city; } public String getCity() { return city; } }

We can now bind this bean to a Form using the BeanItem adapter as follows. // Create a form and use FormLayout as its layout. final Form form = new Form(); // Set form caption and description texts form.setCaption("Contact Information"); form.setDescription("Please specify name of the person and the city where the person lives in."); // Create the custom bean. PersonBean bean = new PersonBean(); // Create a bean item that is bound to the bean. BeanItem item = new BeanItem(bean); // Bind the bean item as the data source for the form. form.setItemDataSource(item);

The Form uses FormLayout layout by default and automatically generates the fields for each of the bean properties, as shown in Figure 5.52, â&#x20AC;&#x153;Form Automatically Generated from a Beanâ&#x20AC;? below.

Figure 5.52. Form Automatically Generated from a Bean

The automatically determined order of the fields can be undesirable. To set the order properly, you can use the setVisibleItemProperties() method of the Form, which takes an ordered collection as its parameter. Fields that are not listed in the collection are not included in the form. // Set the order of the items in the form. Vector order = new Vector(); order.add("city"); order.add("name"); form.setVisibleItemProperties(order);

The form uses the property identifiers as the captions of the fields by default. If you want to have more proper captions for the fields, which is often the case, you need to use a FieldFactory to create the fields, as is shown in the section below.

Generating Proper Fields with a FormFieldFactory The form generates the fields automatically using very coarse logic. A String, int, or double will result in a TextField alike, regardless of the meaning of the field. You might want to have a city

134

Binding Form to Data


User Interface Components

name to be input with a combo box, for example. You can create such custom fields by implementing the createField() method in the FormFieldFactory interface. The default implementation, DefaultFieldFactory is shared with the Table component: it also implements the TableFieldFactory interface. This allows the DefaultFieldFactory to create the fields for both purposes with the same logic. It is usually simplest to just extend the default implementation instead of implementing the interfaces from scratch. You should consult the source code of DefaultFieldFactory to see how it works; you may want to reimplement createFieldByPropertyType(), which actually creates the fields by type, instead of the createField(). Below is an example of implementing the FormFieldFactory interface for a specific form, using the names of the fields of the form to create the editable field components. class MyFieldFactory implements FormFieldFactory { public Field createField(Item item, Object propertyId, Component uiContext) { // Identify the fields by their Property ID. String pid = (String) propertyId; if (pid.equals("name")) { return new TextField("Name"); } else if (pid.equals("city")) { Select select = new Select("City"); select.addItem("Berlin"); select.addItem("Helsinki"); select.addItem("London"); select.addItem("New York"); select.addItem("Turku"); select.setNewItemsAllowed(true); return select; } return null; // Invalid field (property) name. } }

You set a Form to use a custom field factory with setFieldFactory(): form.setFieldFactory(new MyFieldFactory());

The earlier example will now look as shown in Figure 5.53, â&#x20AC;&#x153;Form Fields Generated with a FormFieldFactoryâ&#x20AC;?.

Binding Form to Data

135


User Interface Components

Figure 5.53. Form Fields Generated with a FormFieldFactory

5.17.3. Validating Form Input Validation of the form input is one of the most important tasks in handling forms. The fields in Vaadin can be bound to validators. The validation provides feedback about bad input and the forms can also manage validation results and accept the input only if all validations are successful. Fields can also be set as required, which is a special built-in validator. The validators work on the server-side.

Using Validators in Forms Validators check the validity of input and, if the input is invalid, can provide an error message through an exception. Validators are classes that implement the Validator interface. The interface has two methods that you must implement: isValid() that returns the success or failure as a truth value, and validate(), which reports a failure with an exception. The exception can be associated with an error message describing the details of the error. Simple validators that only need to produce a single error message in case the validation fails can inherit from AbstractValidator or AbstractStringValidator. The Vaadin also provides a number of standard validators, including IntegerValidator and DoubleValidator for validating numerical input, StringLengthValidator, EmailValidator and the more general RegexpValidator for checking that a string matches a Java regular expression: // Postal code that must be 5 digits (10000-99999). TextField field = new TextField("Postal Code"); field.setColumns(5); // Create the validator Validator postalCodeValidator = new RegexpValidator( "[1-9][0-9]{4}", "Postal code must be a number 10000-99999."); field.addValidator(postalCodeValidator);

If you are using a custom FieldFactory to generate the fields, you may want to set the validators for fields there. It is useful to have the form in immediate mode: // Set the form to act immediately on user input. This is // necessary for the validation of the fields to occur immediately // when the input focus changes and not just on commit. form.setImmediate(true);

136

Validating Form Input


User Interface Components

Validation is done always when you call the commit() method of the Form. // The Commit button calls form.commit(). Button commit = new Button("Commit", form, "commit");

If any of the validators in the form fail, the commit will fail and a validation exception message is displayed in the error indicator of the form. If the commit is successful, the input data is written to the data source. Notice that commit() also implicitly sets setValidationVisible(true) (if setValidationVisibleOnCommit() is true, as is the default). This makes the error indicators visible even if they were previously not visible.

Figure 5.54. Form Validation in Action

For cases in which more complex error handling is required, the validator can also implement the Validator interface directly: // Create the validator Validator postalCodeValidator = new Validator() { // The isValid() method returns simply a boolean value, so // it can not return an error message. public boolean isValid(Object value) { if (value == null || !(value instanceof String)) { return false; } return ((String) value).matches("[1-9][0-9]{4}"); } // Upon failure, the validate() method throws an exception // with an error message. public void validate(Object value) throws InvalidValueException { if (!isValid(value)) { if (value != null && value.toString().startsWith("0")) { throw new InvalidValueException( "Postal code must not start with a zero."); } else { throw new InvalidValueException( "Postal code must be a number 10000-99999."); } } } };

Required Fields in Forms Setting a field as required outside a form is usually just a visual clue to the user. Leaving a required field empty does not display any error indicator in the empty field as a failed validation does. However, if you set a form field as required with setRequired(true) and give an error

Validating Form Input

137


User Interface Components

message with setRequiredError() and the user leaves the required field empty, the form will display the error message in its error indicator. form.getField("name").setRequired(true); form.getField("name").setRequiredError("Name is missing"); form.getField("address").setRequired(true); // No error message

To have the validation done immediately when the fields lose focus, you should set the form as immediate, as was done in the section above.

Figure 5.55. Empty Required Field After Clicking Commit

Note It is important that you provide the user with feedback from failed validation of required fields either by setting an error message or by providing the feedback by other means. Otherwise, when a user clicks the Ok button (commits the form), the button does not appear to work and the form does not indicate any reason. As an alternative to setting the error message, you can handle the validation error and provide the feedback about the problem with a different mechanism.

5.17.4. Buffering Form Data Buffering means keeping the edited data in a buffer and writing it to the data source only when the commit() method is called for the component. If the user has made changes to a buffer, calling discard() restores the buffer from the data source. Buffering is actually a feature of all Field components and Form is a Field. Form manages the buffering of its contained fields so that if commit() or discard() is called for the Form, it calls the respective method for all of its managed fields. final Form form = new Form(); ...add components... // Enable buffering. form.setWriteThrough(false); // The Ok button calls form.commit(). Button commit = new Button("Ok", form, "commit"); // The Restore button calls form.discard(). Button restore = new Button("Restore", form, "discard");

The Form example in the Feature Browser of Vaadin demonstrates buffering in forms. The Widget caching demo in Additional demos demonstrates buffering in other Field components, its source code is available in BufferedComponents.java.

138

Buffering Form Data


User Interface Components

5.18. ProgressIndicator The ProgressIndicator component allows displaying the progress of a task graphically. The progress is given as a floating-point value between 0.0 and 1.0.

Figure 5.56. The Progress Indicator Component

The progress indicator polls the server for updates for its value. If the value has changed, the progress is updated. Notice that the user application does not have to handle any polling event, but updating the component is done automatically. Creating a progress indicator is just like with any other component.You can give the initial progress value as a parameter for the constructor. The default polling frequency is 1000 milliseconds (one second), but you can set some other interval with the setPollingInterval() method. // Create the indicator final ProgressIndicator indicator = new ProgressIndicator(new Float(0.0)); main.addComponent(indicator); // Set polling frequency to 0.5 seconds. indicator.setPollingInterval(500);

CSS Style Rules /* Base element. */ .v-progressindicator {} /* Progress indication element on top of the base. */ .v-progressindicator div {}

The default style for the progress indicator uses an animated GIF image (img/base.gif) as the base background for the component. The progress is a <div> element inside the base. When the progress element grows, it covers more and more of the base background. By default, the graphic of the progress element is defined in img/progress.png under the default style directory. S e e com.vaadin.terminal.gwt/public/default/progressindicator/progressindicator.css.

5.18.1. Doing Heavy Computation The progress indicator is often used to display the progress of a heavy server-side computation task. In the following example, we create a thread in the server to do some "heavy work". All the thread needs to do is to set the value of the progress indicator with setValue() and the current progress is displayed automatically when the browser polls the server. // Create an indicator that makes you look busy final ProgressIndicator indicator = new ProgressIndicator(new Float(0.0)); main.addComponent(indicator); // Set polling frequency to 0.5 seconds. indicator.setPollingInterval(500); // Add a button to start working final Button button = new Button("Click to start"); main.addComponent(button);

ProgressIndicator

139


User Interface Components

// Another thread to do some work class WorkThread extends Thread { public void run () { double current = 0.0; while (true) { // Do some "heavy work" try { sleep(50); // Sleep for 50 milliseconds } catch (InterruptedException) {} // Show that you have made // grow the progress value current += 0.01; if (current>1.0) indicator.setValue(new else indicator.setValue(new

some progress: until it reaches 1.0.

Float(1.0)); Float(current));

// After all the "work" has been done for a while, // take a break. if (current > 1.2) { // Restore the state to initial. indicator.setValue(new Float(0.0)); button.setVisible(true); break; } } } } // Clicking the button creates and runs a work thread button.addListener(new Button.ClickListener() { public void buttonClick(ClickEvent event) { final WorkThread thread = new WorkThread(); thread.start(); // The button hides until the work is done. button.setVisible(false); } });

Figure 5.57. Starting Heavy Work

5.19. Slider The Slider is a vertical or horizontal bar that allows setting a numeric value within a defined range by dragging a bar handle with the mouse. The value is shown when dragging the handle. Slider has a number of different constructors that take a combination of the caption, minimum and maximum value, resolution, and the orientation of the slider. // Create a vertical slider final Slider vertslider = new Slider(1, 100); vertslider.setOrientation(Slider.ORIENTATION_VERTICAL);

Slider Properties min

140

Minimum value of the slider range. The default is 0.0.

Slider


User Interface Components

max

Maximum value of the slider range. The default is 100.0.

resolution

The number of digits after the decimal point. The default is 0.

orientation

The orientation can be either horizontal (Slider.ORIENTATION_HORIZONTAL) or vertical (Slider.ORIENTATION_VERTICAL). The default is horizontal.

As the Slider is a field component, you can handle value changes with a ValueChangeListener. The value of the Slider field is a Double object. // Shows the value of the vertical slider final Label vertvalue = new Label(); vertvalue.setSizeUndefined(); // Handle changes in slider value. vertslider.addListener(new Property.ValueChangeListener() { public void valueChange(ValueChangeEvent event) { double value = (Double) vertslider.getValue(); // Use the value box.setHeight((float) value, Sizeable.UNITS_PERCENTAGE); vertvalue.setValue(String.valueOf(value)); } }); // The slider has to be immediate to send the changes // immediately after the user drags the handle. vertslider.setImmediate(true);

You can set the value with the setValue() method defined in Slider that takes the value as a native double value. The setter can throw a ValueOutOfBoundsException, which you must handle. // Set the initial value. This has to be set after the // listener is added if we want the listener to handle // also this value change. try { vertslider.setValue(50.0); } catch (ValueOutOfBoundsException e) { }

Alternatively, you can use the regular setValue(Object), which does not do bounds checking. Figure 5.58, â&#x20AC;&#x153;The Slider Componentâ&#x20AC;? shows both vertical (from the code examples) and horizontal sliders that control the size of a box. The slider values are displayed also in separate labels.

Slider

141


User Interface Components

Figure 5.58. The Slider Component

CSS Style Rules .v-slider {} .v-slider-base {} .v-slider-handle {}

The enclosing style for the Slider is v-slider. The slider bar has style v-slider-base. Even though the handle is higher (for horizontal slider) or wider (for vertical slider) than the bar, the handle element is nevertheless contained within the slider bar element. The appearance of the handle comes from a background image defined in the background CSS property.

5.20. LoginForm The LoginForm component is a login form that allows a password manager in the web browser to remember and later automatically fill in the username and password. This commonly used functionality does not work with regular Vaadin components and is a common problem with Ajax applications. // A wrapper with a caption for the login form Panel loginPanel = new Panel("Login"); loginPanel.setWidth("250px"); LoginForm login = new LoginForm(); loginPanel.addComponent(login);

The resulting form is shown in Figure 5.59, â&#x20AC;&#x153;The LoginForm Componentâ&#x20AC;?.

142

CSS Style Rules


User Interface Components

Figure 5.59. The LoginForm Component

The LoginForm uses static HTML inside an iframe element to enable the functionality. The component provides a default implementation of the static HTML; if you want to change the layout, you need to reimplement the getLoginHtml() method. The login form expands to fill the containing layout as with setSizeFull(); the containing layout must not have an undefined size in either dimension. How the actual form uses this space depends on the static HTML; the iframe element in which the form is contained has fixed size.

5.20.1. Customizing LoginForm Customization of the login form is necessary, for example, if you need to change the layout or internationalize the form. Customization is done by overriding the getLoginHtml() method, which returns the static HTML of the form. The customization layer is very "unvaadin"-like, and at best hack-ish, but dictated by the form management in browsers. Let us look at a custom login form that lets the user of the form to give the field captions: class MyLoginForm extends LoginForm { String usernameCaption; String passwordCaption; String submitCaption; public MyLoginForm(String usernameCaption, String passwordCaption, String submitCaption) { this.usernameCaption = usernameCaption; this.passwordCaption = passwordCaption; this.submitCaption = submitCaption; }

Then we override the method that generates the static HTML for the form: @Override protected byte[] getLoginHTML() { // Application URI needed for submitting form String appUri = getApplication().getURL().toString() + getWindow().getName() + "/"; String x, h, b; // XML header, HTML head and body

The XML header is needed for the validity of the XHTML page: x = "<!DOCTYPE html PUBLIC \"-//W3C//DTD " + "XHTML 1.0 Transitional//EN\" "

Customizing LoginForm

143


User Interface Components

+ "\"http://www.w3.org/TR/xhtml1/" + "DTD/xhtml1-transitional.dtd\">\n";

Notice that it is important to have a newline (\n) at the end of the XML header line. The HTML header part contains JavaScript definitions that handle submitting the form data. It also copies the style sheet references from the parent window. h = + + + + + + + + + + + + + + + + + + + + + + + +

"<head><script type='text/javascript'>" "var setTarget = function() {" " var uri = '" + appUri + "loginHandler';" " var f = document.getElementById('loginf');" " document.forms[0].action = uri;" " document.forms[0].username.focus();" "};" "" "var styles = window.parent.document.styleSheets;" "for(var j = 0; j < styles.length; j++) {\n" " if(styles[j].href) {" " var stylesheet = document.createElement('link');\n" " stylesheet.setAttribute('rel', 'stylesheet');\n" " stylesheet.setAttribute('type', 'text/css');\n" " stylesheet.setAttribute('href', styles[j].href);\n" " document.getElementsByTagName('head')[0]" " .appendChild(stylesheet);\n" " }" "}\n" "function submitOnEnter(e) {" " var keycode = e.keyCode || e.which;" " if (keycode == 13) {document.forms[0].submit();}" "}\n" "</script>" "</head>";

The HTML body element contains the actual form. Notice that it is contained within an inner iframe. The form and the button must have JavaScript calls to submit the form content. b = + + + + + + + + + + + + + + + + + + + + + + + + + +

"<body onload='setTarget();'" " style='margin:0;padding:0; background:transparent;'" " class='" ApplicationConnection.GENERATED_BODY_CLASSNAME + "'>" "<div class='v-app v-app-loginpage'" " style='background:transparent;'>" "<iframe name='logintarget' style='width:0;height:0;" "border:0;margin:0;padding:0;'></iframe>" "<form id='loginf' target='logintarget'" " onkeypress='submitOnEnter(event)'" " method='post'>" "<table>" "<tr><td>" + usernameCaption + "</td>" "<td><input class='v-textfield' style='display:block;'" " type='text' name='username'></td></tr>" "<tr><td>" + passwordCaption + "</td>" " <td><input class='v-textfield'" " style='display:block;' type='password'" " name='password'></td></tr>" "</table>" "<div>" "<div onclick='document.forms[0].submit();'" " tabindex='0' class='v-button' role='button'>" "<span class='v-button-wrap'>" "<span class='v-button-caption'>" submitCaption + "</span>" "</span></div></div></form></div></body>";

Then combine and return the page as a byte array.

144

Customizing LoginForm


User Interface Components

return (x + "<html>" + h + b + "</html>").getBytes(); }

We can use the custom login form as follows: MyLoginForm loginForm = new MyLoginForm("Name of the User", "A passing word", "Login Me Now");

The customized LoginForm is shown in Figure 5.60, â&#x20AC;&#x153;Customizing the LoginFormâ&#x20AC;?.

Figure 5.60. Customizing the LoginForm

Styling with CSS .v-customcomponent {} .v-customcomponent .v-embedded {} .v-app-loginpage {} .v-app-loginpage .v-textfield {} .v-app-loginpage .v-button {}

The LoginForm component is a purely server-side component that extends CustomComponent and therefore has a v-customcomponent base style. If you wish to do any styling for the component, you should give it a custom style name to distinguish it from the regular CustomComponent. The component contains an iframe in an element with v-embedded style. The other styles are defined in the static HTML code returned by the getLoginHTML() method. The default implementation reuses the styles of the TextField and Button components for the input fields and the button, that is, v-textfield and v-button. The root element has the same v-app style as a regular Vaadin application would have, and an additional v-app-loginpage style. ... + "<div class='v-app v-app-loginpage' style=\"background:transparent;\">" ... + "<input class='v-textfield' ... ... + "<div><input class='v-textfield' ... ... <div ... class='v-button' role='button'>

5.21. Component Composition with CustomComponent The ease of making new user interface components is one of the core features of Vaadin. Typically, you simply combine existing built-in components to produce composite components. In many applications, such composite components make up the majority of the user interface. To create a composite component, you need to inherit the CustomComponent and call the setCompositionRoot() in the constructor to set the composition root component. The root component is typically a layout component that contains multiple components. For example:

Styling with CSS

145


User Interface Components

class MyComposite extends CustomComponent { public MyComposite(String message) { // A layout structure used for composition Panel panel = new Panel("My Custom Component"); panel.setContent(new VerticalLayout()); // Compose from multiple components Label label = new Label(message); label.setSizeUndefined(); // Shrink panel.addComponent(label); panel.addComponent(new Button("Ok")); // Set the size as undefined at all levels panel.getContent().setSizeUndefined(); panel.setSizeUndefined(); setSizeUndefined(); // The composition root MUST be set setCompositionRoot(panel); } }

Take note of the sizing when trying to make a customcomponent that shrinks to fit the contained components. You have to set the size as undefined at all levels; the sizing of the composite component and the composition root are separate. You can use the component as follows: MyComposite mycomposite = new MyComposite("Hello");

The rendered component is shown in Figure 5.61, â&#x20AC;&#x153;A Custom Composite Componentâ&#x20AC;?.

Figure 5.61. A Custom Composite Component

You can also inherit any other components, such as layouts, to attain similar composition. Even further, you can create entirely new low-level components, by integrating custom Google Web Toolkit components or by extending the client-side functionality of built-in components. Development of custom GWT components is covered in Chapter 10, Developing Custom Components.

146

Component Composition with CustomComponent


Chapter 6

Managing Layout 6.1. Overview ................................................................................................ 148 6.2. Window and Panel Root Layout ............................................................. 150 6.3. VerticalLayout and HorizontalLayout ................................................. 150 6.4. GridLayout ............................................................................................ 154 6.5. FormLayout .......................................................................................... 158 6.6. Panel ..................................................................................................... 159 6.7. SplitPanel ............................................................................................. 161 6.8. TabSheet ............................................................................................... 163 6.9. Accordion ............................................................................................. 166 6.10. AbsoluteLayout .................................................................................. 167 6.11. CssLayout .......................................................................................... 170 6.12. Layout Formatting ................................................................................ 172 6.13. Custom Layouts ................................................................................... 178 Ever since the ancient xeroxians invented graphical user interfaces, programmers have wanted to make GUI programming ever easier for themselves. Solutions started simple. When GUIs appeared on PC desktops, practically all screens were of the VGA type and fixed into 640x480 size. Mac or X Window System on UNIX were not much different. Everyone was so happy with such awesome graphics resolutions that they never thought that an application would have to work on a radically different screen size. At worst, screens could only grow, they thought, giving more space for more windows. In the 80s, the idea of having a computer screen in your pocket was simply not realistic. Hence, the GUI APIs allowed placing UI components using screen coordinates. Visual Basic and some other systems provided an easy way for the designer to drag and drop components on a fixed-sized window. One would have thought that at least translators would have complained about the awkwardness of such a solution, but apparently they were not,

Book of Vaadin

147


Managing Layout

as non-engineers, heard or at least cared about. At best, engineers could throw at them a resource editor that would allow them to resize the UI components by hand. Such was the spirit back then. After the web was born, layout design was doomed to change for ever. At first, layout didn't matter much, as everyone was happy with plain headings, paragraphs, and a few hyperlinks here and there. Designers of HTML wanted the pages to run on any screen size. The screen size was actually not pixels but rows and columns of characters, as the baby web was really just hypertext, not graphics. That was soon to be changed. The first GUI-based browser, NCSA Mosaic, launched a revolution that culminated in Netscape Navigator. Suddenly, people who had previously been doing advertisement brochures started writing HTML. This meant that layout design had to be easy not just for programmers, but also allow the graphics designer to do his or her job without having to know a thing about programming. The W3C committee designing web standards came up with the CSS (Cascading Style Sheet) specification, which allowed trivial separation of appearance from content. Later versions of HTML followed, XHTML appeared, as did countless other standards. Page description and markup languages are a wonderful solution for static presentations, such as books and most web pages. Real applications, however, need to have more control. They need to be able to change the state of user interface components and even their layout on the run. This creates a need to separate the presentation from content on exactly the right level. Thanks to the attack of graphics designers, desktop applications were, when it comes to appearance, far behind web design. Sun Microsystems had come in 1995 with a new programming language, Java, for writing cross-platform desktop applications. Java's original graphical user interface toolkit, AWT (Abstract Windowing Toolkit), was designed to work on multiple operating systems as well as embedded in web browsers. One of the special aspects of AWT was the layout manager, which allowed user interface components to be flexible, growing and shrinking as needed. This made it possible for the user to resize the windows of an application flexibly and also served the needs of localization, as text strings were not limited to some fixed size in pixels. It became even possible to resize the pixel size of fonts, and the rest of the layout adapted to the new size. Layout management of Vaadin is a direct successor of the web-based concept for separation of content and appearance and of the Java AWT solution for binding the layout and user interface components into objects in programs. Vaadin layout components allow you to position your UI components on the screen in a hierarchical fashion, much like in conventional Java UI toolkits such as AWT, Swing, or SWT. In addition, you can approach the layout from the direction of the web with the CustomLayout component, which you can use to write your layout as a template in XHTML that provides locations of any contained components. The moral of the story is that, because Vaadin is intended for web applications, appearance is of high importance. The solutions have to be the best of both worlds and satisfy artists of both kind: code and graphics. On the API side, the layout is controlled by UI components, particularly the layout components. On the visual side, it is controlled by themes. Themes can contain any HTML, CSS, and JavaScript that you or your web artists create to make people feel good about your software.

6.1. Overview The user interface components in Vaadin can roughly be divided in two groups: components that the user can interact with and layout components for placing the other components to specific places in the user interface. The layout components are identical in their purpose to layout managers in regular desktop frameworks for Java and you can use plain Java to accomplish sophisticated component layouting.

148

Overview


Managing Layout

You start by creating a root layout for the main window, unless you use the default, and then add the other layout components hierarchically, and finally the interaction components as the leaves of the component tree. // Create the main window. Window main = new Window("My Application"); setMainWindow(main); // Set the root layout (VerticalLayout is actually the default). VerticalLayout root = new VerticalLayout(); main.setContent(root); // Add the topmost component. root.addComponent(new Label("The Ultimate Cat Finder")); // Add a horizontal layout for the bottom part. HorizontalLayout bottom = new HorizontalLayout(); root.addComponent(bottom); bottom.addComponent(new Tree("Major Planets and Their Moons")); bottom.addComponent(new Panel()); ...

You will usually need to tune the layout components a bit by setting sizes, expansion ratios, alignments, spacings, and so on. The general settings are described in Section 6.12, “Layout Formatting”, while the layout component specific settings are described in connection with the component. Layouts are coupled with themes that specify various layout features, such as backgrounds, borders, text alignment, and so on. Definition and use of themes is described in Chapter 8, Themes You can see the finished version of the above example in Figure 6.1, “Layout Example”.

Figure 6.1. Layout Example

The alternative for using layout components is to use the special CustomLayout that allows using HTML templates. This way, you can let the web page designers take responsibility of component layouting using their own set of tools. What you lose is the ability to manage the layout dynamically.

Overview

149


Managing Layout

The Visual Editor While you can always program the layout by hand, the Vaadin plugin for the Eclipse IDE includes a visual (WYSIWYG) editor that you can use to create user interfaces visually. The editor generates the code that creates the user interface and is useful for rapid application development and prototyping. It is especially helpful when you are still learning the framework, as the generated code, which is designed to be as reusable as possible, also works as an example of how you create user interfaces with Vaadin. You can find more about the editor in Chapter 7, Visual User Interface Design with Eclipse (experimental).

6.2. Window and Panel Root Layout The Window and its superclass Panel have a single root layout component. The component is usually a Layout, but any ComponentContainer is allowed. When you create the components, they create a default root layout, usually VerticalLayout, but you can change it with the setContent() method. Window main = new Window("My Application"); setMainWindow(main); // Set another root layout for the main window. TabSheet tabsheet = new TabSheet(); main.setContent(tabsheet);

The size of the root layout is the default size of the particular layout component, for example, a VerticalLayout has 100% width and undefined height by default. In many applications, you want to use the full area of the browser view. Setting the components contained inside the root layout to full size is not enough, and would actually lead to an invalid state if the height of the root layout is undefined. // This is actually the default. main.setContent(new VerticalLayout()); // Set the size of the root layout to full width and height. main.getContent().setSizeFull(); // Add a title area on top of the screen. This takes just the // vertical space it needs. main.addComponent(new Label("My Application")); // Add a menu-view area that takes rest of the vertical space. HorizontalLayout menuview = new HorizontalLayout(); menuview.setSizeFull(); main.addComponent(menuview);

See Section 6.12.1, â&#x20AC;&#x153;Layout Sizeâ&#x20AC;? for more information about setting layout sizes.

6.3. VerticalLayout and HorizontalLayout VerticalLayout and HorizontalLayout components are containers for laying out components either vertically or horizontally, respectively. Some components, such as Window and Panel, have a VerticalLayout as the root layout, which you can set with setContent(). Typical use of the layouts goes as follows: VerticalLayout vertical = new VerticalLayout (); vertical.addComponent(new TextField("Name")); vertical.addComponent(new TextField("Street address"));

150

Window and Panel Root Layout


Managing Layout

vertical.addComponent(new TextField("Postal code")); main.addComponent(vertical);

The text fields have a label attached, which will by default be placed above the field. The layout will look on screen as follows:

Using HorizontalLayout gives the following layout:

The layouts can have spacing between the horizontal or vertical cells, defined with setSpacing(), as described in Section 6.12.3, “Layout Cell Spacing”.The contained components can be aligned within their cells with setComponentAlignment(), as described in Section 6.12.2, “Layout Cell Alignment”. You can use setWidth() and setHeight() to specify width and height of a component in either fixed units or relatively with a percentage.

6.3.1. Sizing Contained Components The components contained within an ordered layout can be laid out in a number of different ways depending on how you specify their height or width in the primary direction of the layout component.

Figure 6.2. Component Widths in HorizontalLayout

Figure 6.2, “Component Widths in HorizontalLayout” above gives a summary of the sizing options for a HorizontalLayout. Let us break the figure down as follows. If a VerticalLayout has undefined height or HorizontalLayout undefined width, the layout will shrink to fit the contained components so that there is no extra space between them. HorizontalLayout fittingLayout = new HorizontalLayout(); fittingLayout.setWidth(Sizeable.SIZE_UNDEFINED, 0); fittingLayout.addComponent(new Button("Small")); fittingLayout.addComponent(new Button("Medium-sized"));

Sizing Contained Components

151


Managing Layout

fittingLayout.addComponent(new Button("Quite a big component")); parentLayout.addComponent(fittingLayout);

If such a vertical layout continues below the bottom of a window (a Window object), the window will pop up a vertical scroll bar on the right side of the window area. This way, you get a "web page". If you set a HorizontalLayout to a defined size horizontally or a VerticalLayout vertically, and there is space left over from the contained components, the extra space is distributed equally between the component cells. The components are aligned within these cells according to their alignment setting, top left by default, as in the example below. fixedLayout.setWidth("400px");

Using percentual sizes for components contained in a layout requires answering the question, "Percentage of what?" There is no sensible default answer for this question in the current implementation of the layouts, so in practice, you may not define "100%" size alone. Often, you want to have one component that takes all the available space left over from other components.You need to set its size as 100% and set it as expanding with setExpandRatio(). The second parameter for the method is an expansion ratio, which is relevant if there are more than one expanding component, but its value is irrelevant for a single expanding component. HorizontalLayout layout = new HorizontalLayout(); layout.setWidth("400px"); // These buttons take the minimum size. layout.addComponent(new Button("Small")); layout.addComponent(new Button("Medium-sized")); // This button will expand. Button expandButton = new Button("Expanding component"); // Use 100% of the expansion cell's width. expandButton.setWidth("100%"); // The component must be added to layout before setting the ratio. layout.addComponent(expandButton); // Set the component's cell to expand. layout.setExpandRatio(expandButton, 1.0f); parentLayout.addComponent(layout);

Notice that you must call setExpandRatio() after addComponent(), because the layout can not operate on an component that it doesn't (yet) include. A layout that contains components with percentual size must have a defined size! If a layout has undefined size and a contained component has, say, 100% size, the component would fill the space given by the layout, while the layout would shrink to

152

Sizing Contained Components


Managing Layout

fit the space taken by the component, which is a paradox. This requirement holds for height and width separately.The debug mode allows detecting such invalid cases; see Section 11.4.1, â&#x20AC;&#x153;Debug Modeâ&#x20AC;?. If you specify an expand ratio for multiple components, they will all try to use the available space according to the ratio. HorizontalLayout layout = new HorizontalLayout(); layout.setWidth("400px"); // Create three equally expanding components. String[] captions = { "Small", "Medium-sized", "Quite a big component" }; for (int i = 1; i <= 3; i++) { Button button = new Button(captions[i-1]); button.setWidth("100%"); layout.addComponent(button); // Have uniform 1:1:1 expand ratio. layout.setExpandRatio(button, 1.0f); }

As the example used the same ratio for all components, the ones with more content may have the content cut. Below, we use differing ratios: // Expand ratios for the components are 1:2:3. layout.setExpandRatio(button, i * 1.0f);

If the size of the expanding components is defined as a percentage (typically "100%"), the ratio is calculated from the overall space available for the relatively sized components. For example, if you have a 100 pixels wide layout with two cells with 1.0 and 4.0 respective expansion ratios, and both the components in the layout are set as setWidth("100%"), the cells will have respective widths of 20 and 80 pixels, regardless of the minimum size of the components. However, if the size of the contained components is undefined or fixed, the expansion ratio is of the excess available space. In this case, it is the excess space that expands, not the components. for (int i = 1; i <= 3; i++) { // Button with undefined size. Button button = new Button(captions[i - 1]); layout4.addComponent(button); // Expand ratios are 1:2:3. layout4.setExpandRatio(button, i * 1.0f); }

It is not meaningful to combine expanding components with percentually defined size and components with fixed or undefined size. Such combination can lead to a very unexpected size for the percentually sized components. A percentual size of a component defines the size of the component within its cell. Usually, you use "100%", but a smaller percentage or a fixed size (smaller than the cell size) will leave an

Sizing Contained Components

153


Managing Layout

empty space in the cell and align the component within the cell according to its alignment setting, top left by default. HorizontalLayout layout50 = new HorizontalLayout(); layout50.setWidth("400px"); String[] captions1 = { "Small 50%", "Medium 50%", "Quite a big 50%" }; for (int i = 1; i <= 3; i++) { Button button = new Button(captions1[i-1]); button.setWidth("50%"); layout50.addComponent(button); // Expand ratios for the components are 1:2:3. layout50.setExpandRatio(button, i * 1.0f); } parentLayout.addComponent(layout50);

6.4. GridLayout GridLayout container lays components out on a grid, defined by the number of columns and rows. The columns and rows of the grid serve as coordinates that are used for laying out components on the grid. Each component can use multiple cells from the grid, defined as an area (x1,y1,x2,y2), although they typically take up only a single grid cell. The grid layout maintains a cursor for adding components in left-to-right, top-to-bottom order. If the cursor goes past the bottom-right corner, it will automatically extend the grid downwards by adding a new row. The following example demonstrates the use of GridLayout. The addComponent takes a component and optional coordinates. The coordinates can be given for a single cell or for an area in x,y (column,row) order. The coordinate values have a base value of 0. If coordinates are not given, the cursor will be used. // Create a 4 by 4 grid layout. GridLayout grid = new GridLayout(4, 4); grid.addStyleName("example-gridlayout"); // Fill out the first row using the cursor. grid.addComponent(new Button("R/C 1")); for (int i = 0; i < 3; i++) { grid.addComponent(new Button("Col " + (grid.getCursorX() + 1))); } // Fill out the first column using coordinates. for (int i = 1; i < 4; i++) { grid.addComponent(new Button("Row " + i), 0, i); }

154

GridLayout


Managing Layout

// Add some components of various shapes. grid.addComponent(new Button("3x1 button"), 1, 1, 3, 1); grid.addComponent(new Label("1x2 cell"), 1, 2, 1, 3); InlineDateField date = new InlineDateField("A 2x2 date field"); date.setResolution(DateField.RESOLUTION_DAY); grid.addComponent(date, 2, 2, 3, 3);

The resulting layout will look as follows. The borders have been made visible to illustrate the layout cells.

Figure 6.3. The Grid Layout Component

A component to be placed on the grid must not overlap with existing components. A conflict causes throwing a GridLayout.OverlapsException.

6.4.1. Sizing Grid Cells You can define the size of both a grid layout and its components in either fixed or percentual units, or leave the size undefined altogether, as described in Section 5.3.9, “Sizing Components”. Section 6.12.1, “Layout Size” gives an introduction to sizing of layouts. The size of the GridLayout component is undefined by default, so it will shrink to fit the size of the components placed inside it. In most cases, especially if you set a defined size for the layout but do not set the contained components to full size, there will be some unused space. The position of the non-full components within the grid cells will be determined by their alignment. See Section 6.12.2, “Layout Cell Alignment” for details on how to align the components inside the cells. The components contained within a GridLayout layout can be laid out in a number of different ways depending on how you specify their height or width. The layout options are similar to HorizontalLayout and VerticalLayout, as described in Section 6.3, “VerticalLayout and HorizontalLayout”. A layout that contains components with percentual size must have a defined size! If a layout has undefined size and a contained component has, say, 100% size, the component would fill the space given by the layout, while the layout would shrink to fit the space taken by the component, which is a paradox. This requirement holds

Sizing Grid Cells

155


Managing Layout

for height and width separately.The debug mode allows detecting such invalid cases; see Section 11.4.1, â&#x20AC;&#x153;Debug Modeâ&#x20AC;?. Often, you want to have one or more rows or columns that take all the available space left over from non-expanding rows or columns. You need to set the rows or columns as expanding with setRowExpandRatio() and setColumnExpandRatio(). The first parameter for these methods is the index of the row or column to set as expanding. The second parameter for the methods is an expansion ratio, which is relevant if there are more than one expanding row or column, but its value is irrelevant if there is only one. With multiple expanding rows or columns, the ratio parameter sets the relative portion how much a specific row/column will take in relation with the other expanding rows/columns. GridLayout grid = new GridLayout(3,2); // Layout containing relatively sized components must have // a defined size, here is fixed size. grid.setWidth("600px"); grid.setHeight("200px"); // Add some content String labels [] = { "Shrinking column<br/>Shrinking row", "Expanding column (1:)<br/>Shrinking row", "Expanding column (5:)<br/>Shrinking row", "Shrinking column<br/>Expanding row", "Expanding column (1:)<br/>Expanding row", "Expanding column (5:)<br/>Expanding row" }; for (int i=0; i<labels.length; i++) { Label label = new Label(labels[i], Label.CONTENT_XHTML); label.setWidth(null); // Set width as undefined grid.addComponent(label); } // Set different expansion ratios for the two columns grid.setColumnExpandRatio(1, 1); grid.setColumnExpandRatio(2, 5); // Set the bottom row to expand grid.setRowExpandRatio(1, 1); // Align and size the labels. for (int col=0; col<grid.getColumns(); col++) { for (int row=0; row<grid.getRows(); row++) { Component c = grid.getComponent(col, row); grid.setComponentAlignment(c, Alignment.TOP_CENTER); // Make the labels high to illustrate the empty // horizontal space. if (col != 0 || row != 0) c.setHeight("100%"); } }

156

Sizing Grid Cells


Managing Layout

Figure 6.4. Expanding Rows and Columns in GridLayout

If the size of the contained components is undefined or fixed, the expansion ratio is of the excess space, as in Figure 6.4, “Expanding Rows and Columns in GridLayout” (excess horizontal space is shown in white). However, if the size of the all the contained components in the expanding rows or columns is defined as a percentage, the ratio is calculated from the overall space available for the percentually sized components. For example, if we had a 100 pixels wide grid layout with two columns with 1.0 and 4.0 respective expansion ratios, and all the components in the grid were set as setWidth("100%"), the columns would have respective widths of 20 and 80 pixels, regardless of the minimum size of their contained components.

CSS Style Rules .v-gridlayout {} .v-gridlayout-margin {}

The v-gridlayout is the root element of the GridLayout component. The v-gridlayout-margin is a simple element inside it that allows setting a padding between the outer element and the cells. For styling the individual grid cells, you should style the components inserted in the cells. The implementation structure of the grid can change, so depending on it, as is done in the example below, is not generally recommended. Normally, if you want to have, for example, a different color for a certain cell, just make set the component inside it setSizeFull(), and add a style name for it. Sometimes you may need to use a layout component between a cell and its actual component just for styling. The following example shows how to make the grid borders visible, as in Figure 6.4, “Expanding Rows and Columns in GridLayout”. .v-gridlayout-gridexpandratio { background: blue; /* Creates a "border" around the grid. */ margin: 10px; /* Empty space around the layout. */ } /* Add padding through which the background color shows. */ .v-gridlayout-gridexpandratio .v-gridlayout-margin { padding: 2px; } /* Add cell borders and make the cell backgrounds white. * Warning: This depends heavily on the HTML structure. */ .v-gridlayout-gridexpandratio > div > div > div { padding: 2px; /* Layout background will show through. */ background: white; /* The cells will be colored white. */ } /* Components inside the layout are a safe way to style cells. */

CSS Style Rules

157


Managing Layout

.v-gridlayout-gridexpandratio .v-label { text-align: left; background: #ffffc0; /* Pale yellow */ }

You should beware of margin, padding, and border settings in CSS as they can mess up the layout. The dimensions of layouts are calculated in the Client-Side Engine of Vaadin and some settings can interfere with these calculations. For more information, on margins and spacing, see Section 6.12.3, “Layout Cell Spacing” and Section 6.12.4, “Layout Margins”

6.5. FormLayout FormLayout is the default layout of a Form component. It lays the form fields and their captions out in two columns, with optional indicators for required fields and errors that can be shown for each field. A Form handles additional layout elements itself, including a caption, a form description, a form error indicator, a footer that is often used for buttons and a border. For more information on these, see Section 5.17, “Form”. The field captions can have an icon in addition to the text. // A FormLayout used outside the context of a Form FormLayout fl = new FormLayout(); // Make the FormLayout shrink to its contents fl.setSizeUndefined(); TextField tf = new TextField("A Field"); fl.addComponent(tf); // Mark the first field as required tf.setRequired(true); tf.setRequiredError("The Field may not be empty."); TextField tf2 = new TextField("Another Field"); fl.addComponent(tf2); // Set the second field straing to error state with a message. tf2.setComponentError( new UserError("This is the error indicator of a Field."));

The resulting layout will look as follows. The error message shows in a tooptip when you hover the mouse pointer over the error indicator.

Figure 6.5. A FormLayout Layout for Forms

158

FormLayout


Managing Layout

CSS Style Rules .v-formlayout {} .v-formlayout .v-caption {} /* Columns in a field row. */ .v-formlayout-contentcell {} /* Field content. */ .v-formlayout-captioncell {} /* Field caption. */ .v-formlayout-errorcell {} /* Field error indicator. */ /* Overall style of field rows. */ .v-formlayout-row {} .v-formlayout-firstrow {} .v-formlayout-lastrow {} /* Required field indicator. */ .v-formlayout .v-required-field-indicator {} .v-formlayout-captioncell .v-caption .v-required-field-indicator {} /* Error indicator. */ .v-formlayout-cell .v-errorindicator {} .v-formlayout-error-indicator .v-errorindicator {}

The top-level element of FormLayout has the v-formlayout style. The layout is tabular with three columns: the caption column, the error indicator column, and the field column. These can be styled with v-formlayout-captioncell, v-formlayout-errorcell, and v-formlayout-contentcell, respectively. While the error indicator is shown as a dedicated column, the indicator for required fields is currently shown as a part of the caption column. For information on setting margins and spacing, see also Section 6.12.3, “Layout Cell Spacing” and Section 6.12.4, “Layout Margins”.

6.6. Panel Panel is a simple container with a frame and an optional caption. The content area is bound to a an inner layout component for laying out the contained components. The default content layout is a VerticalLayout, but you can change it with the setContent() method to be any class implementing the ComponentContainer interface. The caption can have an icon in addition to the text. // Create a panel with a caption. final Panel panel = new Panel("Contact Information"); panel.addStyleName("panelexample"); // The width of a Panel is 100% by default, make it // shrink to fit the contents. panel.setWidth(Sizeable.SIZE_UNDEFINED, 0); // Create a layout inside the panel final FormLayout form = new FormLayout(); // Have some margin around it. form.setMargin(true); // Add some components form.addComponent(new TextField("Name")); form.addComponent(new TextField("Email")); // Set the layout as the root layout of the panel panel.setContent(form);

CSS Style Rules

159


Managing Layout

The resulting layout is shown in Figure 6.6, “A Panel Layout in Runo Theme” with the Runo theme.

Figure 6.6. A Panel Layout in Runo Theme

See Section 6.2, “Window and Panel Root Layout” for more information about setting the content layout.

CSS Style Rules .v-panel {} .v-panel-caption {} .v-panel-nocaption {} .v-panel-content {} .v-panel-deco {}

The entire panel has v-panel style. A panel consists of three parts: the caption, content, and bottom decorations (shadow).These can be styled with v-panel-caption, v-panel-content, and v-panel-deco, respectively. If the panel has no caption, the caption element will have the style v-panel-nocaption. The built-in light style has no borders or border decorations for the Panel. You enable it simply by adding the light style name for the panel, as is done in the example below. You can also use the PANEL_LIGHT constant defined in BaseTheme class; it is usable in all subthemes. // Have a window with a SplitPanel. final Window window = new Window("Window with a Light Panel"); window.setWidth("400px"); window.setHeight("200px"); final SplitPanel splitter = new SplitPanel(SplitPanel.ORIENTATION_HORIZONTAL); window.setContent(splitter); // Create a panel with a caption. final Panel light = new Panel("Light Panel"); light.setSizeFull(); // The "light" style is a predefined style without borders light.addStyleName(Runo.PANEL_LIGHT); light.addComponent(new Label("The light Panel has no borders.")); light.getLayout().setMargin(true); // The Panel will act as a "caption" of the left panel // in the SplitPanel. splitter.addComponent(light); splitter.setSplitPosition(250, Sizeable.UNITS_PIXELS); main.addWindow(window);

Figure 6.7, “A Panel with Light Style” shows the rendered Panel in the Runo theme.

160

CSS Style Rules


Managing Layout

Figure 6.7. A Panel with Light Style

The light style is typical when using a Panel as the root layout of a window or some similar layout, as in the example above.

6.7. SplitPanel SplitPanel is a two-component container that divides the available space into two areas to accomodate the two components. The split direction is vertical by default, but you can change it with setOrientation(). You can set the two components with the dedicated setFirstComponent() and setSecondComponent() methods, or with the regular addComponent() method. SplitPanel splitpanel = new SplitPanel(); // Set the orientation. splitpanel.setOrientation(SplitPanel.ORIENTATION_HORIZONTAL); // Put two components in the container. splitpanel.setFirstComponent(new Label("Left Panel")); splitpanel.setSecondComponent(new Label("Right Panel"));

A split bar that divides the two panels is enabled by default. The user can drag the bar with mouse to change the split position. To disable the bar, lock the split position with setLocked(true). The following example shows how you can create a layout with two nested SplitPanel components (one of which has a locked split position): // A top-level panel to put everything in. Panel panel = new Panel("Nested SplitPanels"); // Allow it to shrink to the size of the contained SplitPanel. panel.setSizeUndefined(); // Have a vertical SplitPanel as the main component. SplitPanel vertical = new SplitPanel(); panel.addComponent(vertical); // Set the size of the SplitPanel rather than the containing Panel, // because then we know how much space we have for the panels. vertical.setHeight("150px"); vertical.setWidth("250px"); // Set the split position to 50 pixels, which is more than // enough height for the Label in the upper panel. vertical.setSplitPosition(50, SplitPanel.UNITS_PIXELS);

SplitPanel

161


Managing Layout

// Put a label in the upper panel. vertical.addComponent(new Label("The contents of the upper area.")); // Put a horizontal SplitPanel in the lower area. SplitPanel horizontal = new SplitPanel(); horizontal.setOrientation(SplitPanel.ORIENTATION_HORIZONTAL); horizontal.setSplitPosition(65); // percent vertical.addComponent(horizontal); // The lower SplitPanel is locked, so the user cannot move // the split position. horizontal.setLocked(true); // Component in the left panel: horizontal.addComponent(new Label("Lower left area. "+ "The text on the left wraps around as needed.")); // Component in the right panel: horizontal.addComponent(new Label("Lower right area. "+ "The text on the right also wraps around."));

Figure 6.8. A Layout With Nested SplitPanels

CSS Style Rules /* For a horizontal SplitPanel. */ .v-splitpanel-horizontal {} .v-splitpanel-hsplitter {} .v-splitpanel-hsplitter-locked {} /* For a vertical SplitPanel. */ .v-splitpanel-vertical {} .v-splitpanel-vsplitter {} .v-splitpanel-vsplitter-locked {} /* The two container panels. */ .v-splitpanel-first-container {} /* Top or left panel. */ .v-splitpanel-second-container {} /* Bottom or right panel. */

162

CSS Style Rules


Managing Layout

The entire accordion has the style v-splitpanel-horizontal or v-splitpanel-vertical. The split bar or splitter between the two content panels has either the ...-splitter or ...-splitter-locked style, depending on whether its position is locked or not.

6.8. TabSheet The TabSheet is a multicomponent container that allows switching between the components with "tabs". The tabs are organized as a tab bar at the top of the tab sheet. Clicking on a tab opens its contained component in the main display area of the layout. You add new tabs to a tab sheet with the addTab() method. The simple version of the method takes as its parameter the root component of the tab.You can use the root component to retrieve its corresponding Tab object. Typically, you put a layout component as the root component. // Create an empty tab sheet. TabSheet tabsheet = new TabSheet(); // Create a component to put in a tab and put // some content in it. VerticalLayout myTabRoot = new VerticalLayout(); myTabRoot.addComponent(new Label("Hello, I am a Tab!")); // Add the component to the tab sheet as a new tab. tabsheet.addTab(myTabRoot); // Get the Tab holding the component and set its caption. tabsheet.getTab(myTabRoot).setCaption("My Tab");

Each tab in a tab sheet is represented as a Tab object, which manages the tab caption, icon, and attributes such as hidden and visible. You can set the caption with setCaption() and the icon with setIcon(). If the component added with addTab() has a caption or icon, it is used as the default for the Tab object. However, changing the attributes of the root component later does not affect the tab, but you must make the setting through the Tab object. The addTab() returns the new Tab object, so you can easily set an attribute using the reference. // Set an attribute using the returned reference tabsheet.addTab(myTab).setCaption("My Tab");

You can also give the caption and the icon as parameters for the addTab() method. The following example demonstrates the creation of a simple tab sheet, where each tab shows a different Label component. The tabs have an icon, which are (in this example) loaded as Java class loader resources from the application. TabSheet tabsheet = new TabSheet(); // Make the tabsheet shrink to fit the contents. tabsheet.setSizeUndefined(); tabsheet.addTab(new Label("Contents of the first tab"), "First Tab", new ClassResource("images/Mercury_small.png", this)); tabsheet.addTab(new Label("Contents of the second tab"), "Second Tab", new ClassResource("images/Venus_small.png", this)); tabsheet.addTab(new Label("Contents of the third tab"), "Third tab", new ClassResource("images/Earth_small.png", this));

TabSheet

163


Managing Layout

Figure 6.9. A Simple TabSheet Layout

The hideTabs() method allows hiding the tab bar entirely.This can be useful in tabbed document interfaces (TDI) when there is only one tab. An individual tab can be made invisible by setting setVisible(false) for the Tab object. A tab can be disabled by setting setEnabled(false). Clicking on a tab selects it. This fires a TabSheet.SelectedTabChangeEvent, which you can handle by implementing the TabSheet.SelectedTabChangeListener interface. The source component of the event, which you can retrieve with getSource() method of the event, will be the TabSheet component. You can find the currently selected tab with getSelectedTab() and select (open) a particular tab programmatically with setSelectedTab(). Notice that also adding the first tab fires the SelectedTabChangeEvent, which may cause problems in your handler if you assume that everything is initialized before the first change event. The example below demonstrates handling TabSheet related events and enabling and disabling tabs. The sort of logic used in the example is useful in sequential user interfaces, often called wizards, where the user goes through the tabs one by one, but can return back if needed. import com.vaadin.ui.*; import com.vaadin.ui.Button.ClickEvent; import com.vaadin.ui.TabSheet.SelectedTabChangeEvent; public class TabSheetExample extends CustomComponent implements Button.ClickListener, TabSheet.SelectedTabChangeListener { TabSheet tabsheet = new TabSheet(); Button tab1 = new Button("Push this button"); Label tab2 = new Label("Contents of Second Tab"); Label tab3 = new Label("Contents of Third Tab"); TabSheetExample() { setCompositionRoot(tabsheet); // Listen for changes in tab selection. tabsheet.addListener(this); // First tab contains a button, for which we // listen button click events. tab1.addListener(this); // This will cause a selectedTabChange() call. tabsheet.addTab(tab1, "First Tab", null); // A tab that is initially invisible. tabsheet.addTab(tab2, "Second Tab", null); tabsheet.getTab(tab2).setVisible(false); // A tab that is initially disabled. tabsheet.addTab(tab3, "Third tab", null); tabsheet.getTab(tab3).setEnabled(false); } public void buttonClick(ClickEvent event) { // Enable the invisible and disabled tabs. tabsheet.getTab(tab2).setVisible(true); tabsheet.getTab(tab3).setEnabled(true); // Change selection automatically to second tab.

164

TabSheet


Managing Layout

tabsheet.setSelectedTab(tab2); } public void selectedTabChange(SelectedTabChangeEvent event) { // Cast to a TabSheet. This isn't really necessary in // this example, as we have only one TabSheet component, // but would be useful if there were multiple TabSheets. final TabSheet source = (TabSheet) event.getSource(); if (source == tabsheet) { // If the first tab was selected. if (source.getSelectedTab() == tab1) { // The 2. and 3. tabs may not have been set yet. if (tabsheet.getTab(tab2) != null && tabsheet.getTab(tab3) != null) { tabsheet.getTab(tab2).setVisible(false); tabsheet.getTab(tab3).setEnabled(false); } } } } }

Figure 6.10. A TabSheet with Hidden and Disabled Tabs

CSS Style Rules .v-tabsheet {} .v-tabsheet-tabs {} .v-tabsheet-content {} .v-tabsheet-deco {} .v-tabsheet-tabcontainer {} .v-tabsheet-tabsheetpanel {} .v-tabsheet-hidetabs {} .v-tabsheet-scroller {} .v-tabsheet-scrollerPrev {} .v-tabsheet-scrollerNext {} .v-tabsheet-scrollerPrev-disabled{} .v-tabsheet-scrollerNext-disabled{} .v-tabsheet-tabitem {} .v-tabsheet-tabitem-selected {} .v-tabsheet-tabitemcell {} .v-tabsheet-tabitemcell-first {} .v-tabsheet-tabs td {} .v-tabsheet-spacertd {}

The entire tabsheet has the v-tabsheet style. A tabsheet consists of three main parts: the tabs on the top, the main content pane, and decorations around the tabsheet. The tabs area at the top can be styled with v-tabsheet-tabs, v-tabsheet-tabcontainer and v-tabsheet-tabitem*.

CSS Style Rules

165


Managing Layout

The style v-tabsheet-spacertd is used for any empty space after the tabs. If the tabsheet has too little space to show all tabs, scroller buttons enable browsing the full tab list. These use the styles v-tabsheet-scroller*. The content area where the tab contents are shown can be styled with v-tabsheet-content, and the surrounding decoration with v-tabsheet-deco.

6.9. Accordion Accordion is a multicomponent container similar to TabSheet, except that the "tabs" are arranged vertically. Clicking on a tab opens its contained component in the space between the tab and the next one. You can use an Accordion identically to a TabSheet, which it actually inherits. See Section 6.8, “TabSheet” for more information. The following example shows how you can create a simple accordion. As the Accordion is rather naked alone, we put it inside a Panel that acts as its caption and provides it a border. // Create the Accordion. Accordion accordion = new Accordion(); // Have it take all space available in the layout. accordion.setSizeFull(); // Some components to put in the Accordion. Label l1 = new Label("There are no previously saved actions."); Label l2 = new Label("There are no saved notes."); Label l3 = new Label("There are currently no issues."); // Add the components as tabs in the Accordion. accordion.addTab(l1, "Saved actions", null); accordion.addTab(l2, "Notes", null); accordion.addTab(l3, "Issues", null); // A container for the Accordion. Panel panel = new Panel("Tasks"); panel.setWidth("300px"); panel.setHeight("300px"); panel.addComponent(accordion); // Trim its layout to allow the Accordion take all space. panel.getLayout().setSizeFull(); panel.getLayout().setMargin(false);

Figure 6.11, “An Accordion” shows what the example would look like with the default theme.

166

Accordion


Managing Layout

Figure 6.11. An Accordion

CSS Style Rules .v-accordion {} .v-accordion-item {} .v-accordion-item-open {} .v-accordion-item-first {} .v-accordion-item-caption {} .v-accordion-item-caption .v-caption {} .v-accordion-item-content {}

The top-level element of Accordion has the v-accordion style. An Accordion consists of a sequence of item elements, each of which has a caption element (the tab) and a content area element. The selected item (tab) has also the v-accordion-open style. The content area is not shown for the closed items.

6.10. AbsoluteLayout AbsoluteLayout allows placing components in arbitrary positions in the layout area. The positions are specified in the addComponent() method with horizontal and vertical coordinates relative to an edge of the layout area. The positions can include a third depth dimension, the z-index, which specifies which components are displayed in front and which behind other components. The positions are specified by a CSS absolute position string, using the left, right, top, bottom, and z-index properties known from CSS. In the following example, we have a 300 by 150 pixels large layout and position a text field 50 pixels from both the left and the top edge: // A 400x250 pixels size layout AbsoluteLayout layout = new AbsoluteLayout(); layout.setWidth("400px"); layout.setHeight("250px"); // A component with coordinates for its top-left corner TextField text = new TextField("Somewhere someplace"); layout.addComponent(text, "left: 50px; top: 50px;");

The left and top specify the distance from the left and top edge, respectively. The right and bottom specify the distances from the right and top edge.

CSS Style Rules

167


Managing Layout

// At the top-left corner Button button = new Button( "left: 0px; top: 0px;"); layout.addComponent(button, "left: 0px; top: 0px;"); // At the bottom-right corner Button buttCorner = new Button( "right: 0px; bottom: 0px;"); layout.addComponent(buttCorner, "right: 0px; bottom: 0px;"); // Relative to the bottom-right corner Button buttBrRelative = new Button( "right: 50px; bottom: 50px;"); layout.addComponent(buttBrRelative, "right: 50px; bottom: 50px;"); // On the bottom, relative to the left side Button buttBottom = new Button( "left: 50px; bottom: 0px;"); layout.addComponent(buttBottom, "left: 50px; bottom: 0px;"); // On the right side, up from the bottom Button buttRight = new Button( "right: 0px; bottom: 100px;"); layout.addComponent(buttRight, "right: 0px; bottom: 100px;");

The result of the above code examples is shown in Figure 6.12, “Components Positioned Relative to Various Edges”.

Figure 6.12. Components Positioned Relative to Various Edges

In the above examples, we had components of undefined size and specified the positions of components by a single pair of coordinates. The other possibility is to specify an area and let the component fill the area by specifying a proportinal size for the component, such as "100%". Normally, you use setSizeFull() to take the entire area given by the layout. // Specify an area that a component should fill Panel panel = new Panel("A Panel filling an area"); panel.setSizeFull(); // Fill the entire given area layout.addComponent(panel, "left: 25px; right: 50px; "+ "top: 100px; bottom: 50px;");

The result is shown in Figure 6.13, “Component Filling an Area Specified by Coordinates”

168

AbsoluteLayout


Managing Layout

Figure 6.13. Component Filling an Area Specified by Coordinates

You can also use proportional coordinates to specify the coordinates: // A panel that takes 30% to 90% horizontally and // 20% to 80% vertically Panel panel = new Panel("A Panel"); panel.setSizeFull(); // Fill the specified area layout.addComponent(panel, "left: 30%; right: 10%;" + "top: 20%; bottom: 20%;");

The result is shown in Figure 6.14, “Specifying an Area by Proportional Coordinates”

Figure 6.14. Specifying an Area by Proportional Coordinates

Drag and drop is very useful for moving the components contained in an AbsoluteLayout. Check out the example in Section 11.13.6, “Dropping on a Component”.

Styling with CSS .v-absolutelayout {} .v-absolutelayout-wrapper {}

The AbsoluteLayout component has v-absolutelayout root style. Each component in the layout is contained within an element that has the v-absolutelayout-wrapper.The component

Styling with CSS

169


Managing Layout

captions are outside the wrapper elements, in a separate element with the usual v-caption style.

6.11. CssLayout CssLayout allows strong control over styling of the components contained inside the layout. The components are contained in a simple DOM structure consisting of <div> elements. By default, the contained components are laid out horizontally and wrap naturally when they reach the width of the layout, but you can control this and most other behaviour with CSS. You can also inject custom CSS for each contained component. As CssLayout has a very simple DOM structure and no dynamic rendering logic, relying purely on the built-in rendering logic of the browsers, it is the fastest of the layout components. The basic use of CssLayout is just like with any other layout component: CssLayout layout = new CssLayout(); // Component with a layout-managed caption and icon TextField tf = new TextField("A TextField"); tf.setIcon(new ThemeResource("icons/user.png")); layout.addComponent(tf); // Labels are 100% wide by default so must unset width Label label = new Label("A Label"); label.setWidth(Sizeable.SIZE_UNDEFINED, 0); layout.addComponent(label); layout.addComponent(new Button("A Button"));

The result is shown in Figure 6.15, â&#x20AC;&#x153;Basic Use of CssLayoutâ&#x20AC;?. Notice that the default spacing and alignment of the layout is quite crude and CSS styling is nearly always needed.

Figure 6.15. Basic Use of CssLayout

The display attribute of CssLayout is inline-block by default, so the components are laid out horizontally following another. CssLayout has 100% width by default. If the components reach the width of the layout, they are wrapped to the next "line" just as text would be. If you add a component with 100% width, it will take an entire line by wrapping before and after the component. Overriding the getCss() method allows injecting custom CSS for each component. The CSS returned by the method is inserted in the style attribute of the <div> element of the component, so it will override any style definitions made in CSS files. CssLayout layout = new CssLayout() { @Override protected String getCss(Component c) { if (c instanceof Label) { // Color the boxes with random colors int rgb = (int) (Math.random()*(1<<24)); return "background: #" + Integer.toHexString(rgb); } return null; } }; layout.setWidth("400px"); // Causes to wrap the contents

170

CssLayout


Managing Layout

// Add boxes of various sizes for (int i=0; i<40; i++) { Label box = new Label("&nbsp;", Label.CONTENT_XHTML); box.addStyleName("flowbox"); box.setWidth((float) Math.random()*50, Sizeable.UNITS_PIXELS); box.setHeight((float) Math.random()*50, Sizeable.UNITS_PIXELS); layout.addComponent(box); }

The style name added to the components allows making common styling in a CSS file: .v-label-flowbox { border: thin black solid; }

Figure 6.16, â&#x20AC;&#x153;Use of getCss() and line wrapâ&#x20AC;? shows the rendered result.

Figure 6.16. Use of getCss() and line wrap

The stregth of the CssLayout is also its weakness. Much of the logic behind the other layout components is there to give nice default behaviour and to handle the differences in different browsers. Some browsers, no need to say which, are notoriously incompatible with the CSS standards, so they require a lot of custom CSS. You may need to make use of the browser-specific style classes in the root element of the application. Some features in the other layouts are not even solvable in pure CSS, at least in all browsers.

Styling with CSS .v-csslayout {} .v-csslayout-margin {} .v-csslayout-container {}

The CssLayout component has v-csslayout root style. The margin element with v-csslayout-margin style is always enabled. The components are contained in an element with v-csslayout-container style. For example, we could style the basic CssLayout example shown earlier as follows: /* Have the caption right of the text box, bottom-aligned */ .csslayoutexample .mylayout .v-csslayout-container { direction: rtl; line-height: 24px; vertical-align: bottom; } /* Have some space before and after the caption */ .csslayoutexample .mylayout .v-csslayout-container .v-caption {

Styling with CSS

171


Managing Layout

padding-left: 3px; padding-right: 10px; }

The example would now be rendered as shown in Figure 6.17, “Styling CssLayout”.

Figure 6.17. Styling CssLayout

Captions and icons that are managed by the layout are contained in an element with v-caption style. These caption elements are contained flat at the same level as the actual component elements. This may cause problems with wrapping in inline-block mode, as wrapping can occur between the caption and its corresponding component element just as well as between components. Such use case is therefore not feasible.

6.12. Layout Formatting While the formatting of layouts is mainly done with style sheets, just as with other components, style sheets are not ideal or even possible to use in some situations. For example, CSS does not allow defining the spacing of table cells, which is done with the cellspacing attribute in HTML. Moreover, as many layout sizes are calculated dynamically in the Client-Side Engine of Vaadin, some CSS settings can fail altogether.

6.12.1. Layout Size The size of a layout component can be specified with the setWidth() and setHeight() methods defined in the Sizeable interface, just like for any component. It can also be undefined, in which case the layout shrinks to fit the component(s) inside it. Section 5.3.9, “Sizing Components” gives details on the interface.

Figure 6.18. HorizontalLayout with Undefined vs Defined size

Many layout components take 100% width by default, while they have the height undefined. The sizes of components inside a layout can also be defined as a percentage of the space available in the layout, for example with setWidth("100%"); or with the (most commonly used method) setFullSize() that sets 100% size in both directions. If you use a percentage in a HorizontalLayout, VerticalLayout, or GridLayout, you will also have to set the component as expanding, as noted below. Warning A layout that contains components with percentual size must have a defined size! If a layout has undefined size and a contained component has, say, 100% size, the component will try to fill the space given by the layout, while the layout will shrink to

172

Layout Formatting


Managing Layout

fit the space taken by the component, which is a paradox. This requirement holds for height and width separately.The debug mode allows detecting such invalid cases; see Section 11.4.1, “Debug Mode”. For example: // This takes 100% width but has undefined height. VerticalLayout layout = new VerticalLayout(); // A button that takes all the space available in the layout. Button button = new Button("100%x100% button"); button.setSizeFull(); layout.addComponent(button); // We must set the layout to a defined height vertically, in // this case 100% of its parent layout, which also must // not have undefined size. layout.setHeight("100%");

The default layout of Window and Panel is VerticalLayout with undefined height. If you insert enough components in such a layout, it will grow outside the bottom of the view area and scrollbars will appear in the browser. If you want your application to use all the browser view, nothing more or less, you should use setFullSize() for the root layout. // Create the main window. Window main = new Window("Main Window"); setMainWindow(main); // Use full size. main.getLayout().setSizeFull();

Expanding Components If you set a HorizontalLayout to a defined size horizontally or a VerticalLayout vertically, and there is space left over from the contained components, the extra space is distributed equally between the component cells. The components are aligned within these cells, according to their alignment setting, top left by default, as in the example below.

Often, you don't want such empty space, but want one or more components to take all the leftover space. You need to set such a component to 100% size and use setExpandRatio(). If there is just one such expanding component in the layout, the ratio parameter is irrelevant.

If you set multiple components as expanding, the expand ratio dictates how large proportion of the available space (overall or excess depending on whether the components are sized as a percentage or not) each component takes. In the example below, the buttons have 1:2:3 ratio for the expansion.

GridLayout has corresponding method for both of its directions, setRowExpandRatio() and setColumnExpandRatio(). Expansion is dealt in detail in the documentation of the layout components that support it. See Section 6.3, “VerticalLayout and HorizontalLayout” and Section 6.4, “GridLayout” for details on components with relative sizes.

Layout Size

173


Managing Layout

6.12.2. Layout Cell Alignment You can set the alignment of the component inside a specific layout cell with the setComponentAlignment() method. The method takes as its parameters the component contained in the cell to be formatted, and the horizontal and vertical alignment. Figure 6.19, â&#x20AC;&#x153;Cell Alignmentsâ&#x20AC;? illustrates the alignment of components within a GridLayout.

Figure 6.19. Cell Alignments

The easiest way to set alignments is to use the constants defined in the Alignment class. Let us look how the buttons in the top row of the above GridLayout are aligned with constants: // Create a grid layout final GridLayout grid = new GridLayout(3, 3); grid.setWidth(400, Sizeable.UNITS_PIXELS); grid.setHeight(200, Sizeable.UNITS_PIXELS); Button topleft = new Button("Top Left"); grid.addComponent(topleft, 0, 0); grid.setComponentAlignment(topleft, Alignment.TOP_LEFT); Button topcenter = new Button("Top Center"); grid.addComponent(topcenter, 1, 0); grid.setComponentAlignment(topcenter, Alignment.TOP_CENTER); Button topright = new Button("Top Right"); grid.addComponent(topright, 2, 0); grid.setComponentAlignment(topright, Alignment.TOP_RIGHT); ...

The following table lists all the Alignment constants by their respective locations:

Table 6.1. Alignment Constants TOP_LEFT

TOP_CENTER

TOP_RIGHT

MIDDLE_LEFT MIDDLE_CENTER MIDDLE_RIGHT BOTTOM_LEFT BOTTOM_CENTER BOTTOM_RIGHT Another way to specify the alignments is to create an Alignment object and specify the horizontal and vertical alignment with separate constants. You can specify either of the directions, in which case the other alignment direction is not modified, or both with a bitmask operation between the two directions. Button middleleft = new Button("Middle Left"); grid.addComponent(middleleft, 0, 1); grid.setComponentAlignment(middleleft, new Alignment(Bits.ALIGNMENT_VERTICAL_CENTER |

174

Layout Cell Alignment


Managing Layout

Bits.ALIGNMENT_LEFT)); Button middlecenter = new Button("Middle Center"); grid.addComponent(middlecenter, 1, 1); grid.setComponentAlignment(middlecenter, new Alignment(Bits.ALIGNMENT_VERTICAL_CENTER | Bits.ALIGNMENT_HORIZONTAL_CENTER)); Button middleright = new Button("Middle Right"); grid.addComponent(middleright, 2, 1); grid.setComponentAlignment(middleright, new Alignment(Bits.ALIGNMENT_VERTICAL_CENTER | Bits.ALIGNMENT_RIGHT));

Obviously, you may combine only one vertical bitmask with one horizontal bitmask, though you may leave either one out. The following table lists the available alignment bitmask constants:

Table 6.2. Alignment Bitmasks Horizontal Bits.ALIGNMENT_LEFT Bits.ALIGNMENT_HORIZONTAL_CENTER Bits.ALIGNMENT_RIGHT Vertical

Bits.ALIGNMENT_TOP Bits.ALIGNMENT_VERTICAL_CENTER Bits.ALIGNMENT_BOTTOM

You can determine the current alignment of a component with getComponentAlignment(), which returns an Alignment object. The class provides a number of getter methods for decoding the alignment, which you can also get as a bitmask value.

6.12.3. Layout Cell Spacing The VerticalLayout, HorizontalLayout, and GridLayout layouts offer a setSpacing() method for enabling space between the cells in the layout. Enabling the spacing adds a spacing style for all cells except the first so that, by setting the left or top padding, you can specify the amount of spacing. To enable spacing, simply call setSpacing(true) for the layout as follows: HorizontalLayout layout2 = new HorizontalLayout(); layout2.addStyleName("spacingexample"); layout2.setSpacing(true); layout2.addComponent(new Button("Component 1")); layout2.addComponent(new Button("Component 2")); layout2.addComponent(new Button("Component 3")); VerticalLayout layout4 = new VerticalLayout(); layout4.addStyleName("spacingexample"); layout4.setSpacing(true); layout4.addComponent(new Button("Component 1")); layout4.addComponent(new Button("Component 2")); layout4.addComponent(new Button("Component 3"));

In practise, the setSpacing() method toggles between the "v-COMPONENTCLASSNAME-spacing-on" and "-off" CSS class names in the cell elements. Elements having those class names can be used to define the spacing metrics in a theme.

Layout Cell Spacing

175


Managing Layout

The layouts have a spacing style name to define spacing also when spacing is off. This allows you to define a small default spacing between components by default and a larger one when the spacing is actually enabled. Spacing can be horizontal (for HorizontalLayout), vertical (for VerticalLayout), or both (for GridLayout). The name of the spacing style for horizontal and vertical spacing is the base name of the component style name plus the "-spacing-on" suffix, as shown in the following table:

Table 6.3. Spacing Style Names VerticalLayout

v-verticallayout-spacing-on

HorizontalLayout v-horizontallayout-spacing-on GridLayout

v-gridlayout-spacing-on

In the CSS example below, we specify the exact amount of spacing for the code example given above, for the layouts with the custom "spacingexample" style: /* Set the amount of horizontal cell spacing in a * specific element with the "-spacingexample" style. */ .v-horizontallayout-spacingexample .v-horizontallayout-spacing-on { padding-left: 30px; } /* Set the amount of vertical cell spacing in a * specific element with the "-spacingexample" style. */ .v-verticallayout-spacingexample .v-verticallayout-spacing-on { padding-top: 30px; } /* Set the amount of both vertical and horizontal cell spacing * in a specific element with the "-spacingexample" style. */ .v-gridlayout-spacingexample .v-gridlayout-spacing-on { padding-top: 30px; padding-left: 50px; }

The resulting layouts will look as shown in Figure 6.20, â&#x20AC;&#x153;Layout Spacingsâ&#x20AC;?, which also shows the layouts with no spacing.

Figure 6.20. Layout Spacings

Note Spacing is unrelated to "cell spacing" in HTML tables. While many layout components are implemented with HTML tables in the browser, this implementation is not guaranteed to stay the same and at least Vertical-/HorizontalLayout could be implemented with <div> elements as well. In fact, as GWT compiles widgets separately for different browsers, the implementation could even vary between browsers.

176

Layout Cell Spacing


Managing Layout

Also note that HTML elements with spacing classnames don't necessarily exist in a component after rendering, because the Client-Side Engine of Vaadin processes them.

6.12.4. Layout Margins By default, layout components do not have any margin around them. You can add margin with CSS directly to the layout component. Below we set margins for a specific layout component (here a horizontallayout): layout1.addStyleName("marginexample1"); .v-horizontallayout-marginexample1 .v-horizontallayout-margin { padding-left: 200px; padding-right: 100px; padding-top: 50px; padding-bottom: 25px; }

Similar settings exist for other layouts such as verticallayout. The layout size calculations require the margins to be defined as CSS padding rather than as CSS margin. As an alternative to the pure CSS method, you can set up a margin around the layout that can be enabled with setMargin(true). The margin element has some default margin widths, but you can adjust the widths in CSS if you need to. Let us consider the following example, where we enable the margin on all sides of the layout: // Create a layout HorizontalLayout layout2 = new HorizontalLayout(); containinglayout.addComponent( new Label("Layout with margin on all sides:")); containinglayout.addComponent(layout2); // Set style name for the layout to allow styling it layout2.addStyleName("marginexample"); // Have margin on all sides around the layout layout2.setMargin(true); // Put something inside the layout layout2.addComponent(new Label("Cell 1")); layout2.addComponent(new Label("Cell 2")); layout2.addComponent(new Label("Cell 3"));

You can enable the margins only for specific sides. The margins are specified for the setMargin() method in clockwise order for top, right, bottom, and left margin. The following would enable the top and left margins: layout2.setMargin(true, false, false, true);

You can specify the actual margin widths in the CSS if you are not satisfied with the default widths (in this example for a HorizontalLayout): .v-horizontallayout-marginexample .v-horizontallayout-margin-left 200px;} .v-horizontallayout-marginexample .v-horizontallayout-margin-right 100px;} .v-horizontallayout-marginexample .v-horizontallayout-margin-top

Layout Margins

{padding-left: {padding-right: {padding-top:

177


Managing Layout

50px; } .v-horizontallayout-marginexample .v-horizontallayout-margin-bottom {padding-bottom: 25px; }

The resulting margins are shown in Figure 6.21, â&#x20AC;&#x153;Layout Marginsâ&#x20AC;? below. The two ways produce identical margins.

Figure 6.21. Layout Margins

CSS Style Rules The CSS style names for the margin widths for setMargin() consist of the specific layout name plus -margin-left and so on. The CSS style names for CSS-only margins consist of the specific layout name plus -margin. Below, the style rules are given for VerticalLayout: /* Alternative 1: CSS only style */ .v-verticallayout-margin { padding-left: ___px; padding-right: ___px; padding-top: ___px; padding-bottom: ___px; } /* Alternative 2: CSS rules to be enabled in code */ .v-verticallayout-margin-left {padding-left: ___px;} .v-verticallayout-margin-right {padding-right: ___px;} .v-verticallayout-margin-top {padding-top: ___px;} .v-verticallayout-margin-bottom {padding-bottom: ___px;}

6.13. Custom Layouts While it is possible to create almost any typical layout with the standard layout components, it is sometimes best to separate the layout completely from code. With the CustomLayout component, you can write your layout as a template in XHTML that provides locations of any contained components. The layout template is included in a theme. This separation allows the layout to be designed separately from code, for example using WYSIWYG web designer tools such as Adobe Dreamweaver. A template is a HTML file located under layouts folder under a theme folder under the WebContent/VAADIN/themes/ fo l d e r, for example, WebContent/VAADIN/themes/themename/layouts/mylayout.html. (Notice that the root path WebContent/VAADIN/themes/ for themes is fixed.) A template can also be provided

178

Custom Layouts


Managing Layout

dynamically from an InputStream, as explained below. A template includes <div> elements with a location attribute that defines the location identifier. All custom layout HTML-files must be saved using UTF-8 character encoding. <table width="100%" height="100%"> <tr height="100%"> <td> <table align="center"> <tr> <td align="right">User&nbsp;name:</td> <td><div location="username"></div></td> </tr> <tr> <td align="right">Password:</td> <td><div location="password"></div></td> </tr> </table> </td> </tr> <tr> <td align="right" colspan="2"> <div location="okbutton"></div> </td> </tr> </table>

The client-side engine of Vaadin will replace contents of the location elements with the components. The components are bound to the location elements by the location identifier given to addComponent(), as shown in the example below. // Have a Panel where to put the custom layout. Panel panel = new Panel("Login"); panel.setSizeUndefined(); main.addComponent(panel); // Create custom layout from "layoutname.html" template. CustomLayout custom = new CustomLayout("layoutname"); custom.addStyleName("customlayoutexample"); // Use it as the layout of the Panel. panel.setContent(custom); // Create a few components and bind them to the location tags // in the custom layout. TextField username = new TextField(); custom.addComponent(username, "username"); TextField password = new TextField(); custom.addComponent(password, "password"); Button ok = new Button("Login"); custom.addComponent(ok, "okbutton");

The resulting layout is shown below in Figure 6.22, â&#x20AC;&#x153;Example of a Custom Layout Componentâ&#x20AC;?.

Custom Layouts

179


Managing Layout

Figure 6.22. Example of a Custom Layout Component

You can use addComponent() also to replace an existing component in the location given in the second parameter. In addition to a static template file, you can provide a template dynamically with the CustomLayout constructor that accepts an InputStream as the template source. For example: new CustomLayout(new ByteArrayInputStream("<b>Template</b>".getBytes()));

or new CustomLayout(new FileInputStream(file));

180

Custom Layouts


Chapter 7

Visual User Interface Design with Eclipse (experimental) 7.1. Overview ................................................................................................ 182 7.2. Creating a New CustomComponent .................................................... 182 7.3. Using The Visual Editor ......................................................................... 184 7.4. Structure of a Visually Editable Component .......................................... 189 This chapter provides instructions for developing the graphical user interface of Vaadin components with a visual editor (experimental) included in the Vaadin Plugin for the Eclipse IDE. A revised visual editor plugin will be published around the time of publication of Vaadin 6.4. This chapter describes the experimental visual editor available in early 2010.

Book of Vaadin

181


Visual User Interface Design with Eclipse (experimental)

7.1. Overview The visual or WYSIWYG editor for Vaadin allows you to design the user interface of an entire application or specific custom components. The editor generates the actual Java code, which is designed to be reusable, so you can design the basic layout of the user interface with the visual editor and build the user interaction logic on top of the generated code. You can use inheritance and composition to modify the components further. The editor is included in the Vaadin Plugin for Eclipse (actually as a separate plugin in the plugin set). For installing the Vaadin plugin, see Section 2.2.5, “Vaadin Plugin for Eclipse”. Ongoing Work The first preview version of the visual editor was released in May 2009 and is still under development at the time of the publication of this book and should be considered as experimental. While the preview version is incomplete and probably not suitable for demanding use, you can use it for simple tasks, especially when familiarizing yourself with Vaadin.

7.2. Creating a New CustomComponent The visual editor works with custom component classes that extend the CustomComponent class, which is the basic technique in Vaadin for creating composite components. Custom components are described in Section 5.21, “Component Composition with CustomComponent”. Any CustomComponent will not do for the visual editor; you need to create a new one as instructed below. 1. Select File → New → Other... in the main menu or right-click the Project Explorer and select New → Other... to open the New window. 2. In the first, Select a wizard step, select Vaadin → Vaadin CustomComponent and click Next.

182

Overview


Visual User Interface Design with Eclipse (experimental)

3. The Source folder is the root source directory where the new component will be created. This is by default the default source directory of your project.

Enter the Java Package under which the new component class should be created or select it by clicking the Browse button. Also enter the class Name of the new component. Finally, click Finish to create the component. A newly created composite component is opened in the editor window, as shown in Figure 7.1, â&#x20AC;&#x153;New Composite Componentâ&#x20AC;?.

Creating a New CustomComponent

183


Visual User Interface Design with Eclipse (experimental)

Figure 7.1. New Composite Component

You can observe that a component that you can edit with the visual editor has two tabs at the bottom of the view: Source and Design. These tabs allow switching between the source view and the visual editing view.

7.3. Using The Visual Editor The visual editor view consists of, on the left side, an editing area that displays the current layout and, on the right side, a control panel that contains a component tree, component property panel, and a component list for adding new components.

7.3.1. Adding New Components Adding new components in the user interface is done by adding nodes to the component tree at the top of the editor control panel. The list box at the bottom of the editor control panel shows the components allowed under the selected component; you can generally add components only under a layout (Layout) or a component container (ComponentContainer). To add a new node:

184

Using The Visual Editor


Visual User Interface Design with Eclipse (experimental)

1. select an existing node (in the component tree) under which you wish to insert the new node 2. click on a component in the component list at the bottom of the panel.

Figure 7.2. Adding a New Component Node

The component will be added under the selected component in the component tree. You can delete a component by right-clicking it in the component tree and selecting Remove.

7.3.2. Setting Component Properties The property setting area of the control panel allows setting component properties. The area is divided into basic properties, size and position settings, and other properties, which includes also styles.

Basic Properties The top section of the property panel, shown in Figure 7.3, â&#x20AC;&#x153;Basic Component Propertiesâ&#x20AC;?, allows settings basic component properties.

Setting Component Properties

185


Visual User Interface Design with Eclipse (experimental)

Figure 7.3. Basic Component Properties

The basic properties are: Component name

The name of the component, which is used for the reference to the component, so it must obey Java notation for variable names.

Caption

The caption of a component is usually displayed above the component. Some components, such as Button, display the caption inside the component. For Label text, you should set the value of the label instead of the caption, which should be left empty.

Description (tooltip)

The description is usually displayed as a tooltip when the mouse pointer hovers over the component for a while. Some components, such as Form have their own way of displaying the description.

Icon

The icon of a component is usually displayed above the component, left of the caption. Some components, such as Button, display the icon inside the component.

Formatting type

Some components allow different formatting types, such as Label, which allow formatting either as Text, XHTML, Preformatted, and Raw.

Value

The component value. The value type and how it is displayed by the component varies between different component types and each value type has its own editor. The editor opens by clicking on the ... button.

Most of the basic component properties are defined in the Component interface; see Section 5.2.1, “Component Interface” for further details.

Size and Position The size of a component is determined by its width and height, which you can give in the two edit boxes in the control panel. You can use any unit specifiers for components, as described in Section 5.3.9, “Sizing Components”. Emptying a size box will make the size "automatic", which means setting the size as undefined. In the generated code, the undefined value will be expressed as "-1px". Setting width of "100px" and auto (undefined or empty) height would result in the following generated settings for a button:

186

Setting Component Properties


Visual User Interface Design with Eclipse (experimental)

// myButton myButton = new Button(); ... myButton.setHeight("-1px"); myButton.setWidth("100px"); ...

Figure 7.4, “Component Size and Position” shows the control panel area for the size and position.

Figure 7.4. Component Size and Position

The generated code for the example would be: // myButton myButton = new Button(); myButton.setWidth("-1px"); myButton.setHeight("-1px"); myButton.setImmediate(true); myButton.setCaption("My Button"); mainLayout.addComponent(myButton, "top:243.0px;left:152.0px;");

The position is given as a CSS position in the second parameter for addComponent(). The values "-1px" for width and height will make the button to be sized automatically to the minimum size required by the caption. When editing the position of a component inside an AbsoluteLayout, the editor will display vertical and horizontal guides, which you can use to set the position of the component. See Section 7.3.3, “Editing an AbsoluteLayout” for more information about editing absolute layouts. The ZIndex setting controls the "Z coordinate" of the components, that is, which component will overlay which when they overlap. Value -1 means automatic, in which case the components added to the layout later will be on top.

Other Properties The bottom section of the property panel, shown in Figure 7.5, “Other Properties”, allows settings other often used component properties.

Figure 7.5. Other Properties

Setting Component Properties

187


Visual User Interface Design with Eclipse (experimental)

Style Names

Enter the enabled CSS style names for the component as a spaceseraparated list. See Chapter 8, Themes for information on component styles in themes.

Visible

Defines whether the component is visible or hidden, as set with the setVisible() method.

Enabled

Defines whether the component is enabled or disabled, as set with the setEnabled() method. If disabled, no user interaction is allowed for the component and it is usually shown as shadowed.

Read-only

Defines whether editing the component is allowed, as set with the setReadonly() method.

Immediate

Defines whether user interaction with the component is updated immediately (after the component loses focus) in the server as a state or value chage in the component value.

Most of the component properties mentioned above are defined in the Component interface; see Section 5.2.1, “Component Interface” for further details.

7.3.3. Editing an AbsoluteLayout The visual editor has interactive support for the AbsoluteLayout component that allows positioning components exactly at specified coordinates.You can position the components using guides that control the position attributes, shown in the control panel on the right. The position values are measured in pixels from the corresponding edge; the vertical and horizontal rulers show the distances from the top and left edge. Figure 7.6, “Positioning with AbsoluteLayout” shows three components, a Label, a Table, and a Button, inside an AbsoluteLayout.

Figure 7.6. Positioning with AbsoluteLayout

188

Editing an AbsoluteLayout


Visual User Interface Design with Eclipse (experimental)

Position attributes that are empty are automatic and can be either zero (at the edge) or dynamic to make it shrink to fit the size of the component, depending on the component. Guides are shown also for the automatic position attributes and move automatically; in Figure 7.6, “Positioning with AbsoluteLayout” the right and bottom edges of the Button are automatic. Moving an automatic guide manually makes the guide and the corresponding the position attribute non-automatic. To make a manually set attribute automatic, empty it in the control panel. Figure 7.7, “Manually positioned Label” shows a Label component with all the four edges set manually. Notice that if an automatic position is 0, the guide is at the edge of the ruler.

Figure 7.7. Manually positioned Label

7.4. Structure of a Visually Editable Component A component created by the wizard and later managed by the visual editor has a very specific structure that allows you to insert your user interface logic in the component while keeping a minimal amount of code off-limits. You need to know what you can edit yourself and what exactly is managed by the editor. The managed member variables and methods are marked with the AutoGenerated annotation, as you can see later. A visually editable component consists of: • Member variables containing sub-component references • Sub-component builder methods • The constructor The structure of a composite component is hierarchical, a nested hierarchy of layout components containing other layout components as well as regular components. The root layout of the component tree, or the composition root of the CustomComponent, is named mainLayout. See Section 5.21, “Component Composition with CustomComponent” for a detailed description of the structure of custom (composite) components.

Structure of a Visually Editable Component

189


Visual User Interface Design with Eclipse (experimental)

7.4.1. Sub-Component References The CustomComponent class will include a reference to each contained component as a member variable. The most important of these is the mainLayout reference to the composition root layout. Such automatically generated member variables are marked with the @AutoGenerated annotation. They are managed by the editor, so you should not edit them manually, unless you know what you are doing. A composite component with an AbsoluteLayout as the composition root, containing a Button and a Table would have the references as follows: public class MyComponent extends CustomComponent { @AutoGenerated private AbsoluteLayout mainLayout; @AutoGenerated private Button myButton; @AutoGenerated private Table myTable; ...

The names of the member variables are defined in the component properties panel of the visual editor, in the Component name field, as described in the section called “Basic Properties”. While you can change the name of any other components, the name of the root layout is always mainLayout. It is fixed because the editor does not make changes to the constructor, as noted in Section 7.4.3, “The Constructor”. You can, however, change the type of the root layout, which is an AbsoluteLayout by default. Certain typically static components, such as the Label label component, will not have a reference as a member variable. See the description of the builder methods below for details.

7.4.2. Sub-Component Builders Every managed layout component will have a builder method that creates the layout and all its contained components. The builder puts references to the created components in their corresponding member variables, and it also returns a reference to the created layout component. Below is an example of an initial main layout: @AutoGenerated private AbsoluteLayout buildMainLayout() { // common part: create layout mainLayout = new AbsoluteLayout(); // top-level component properties setHeight("100.0%"); setWidth("100.0%"); return mainLayout; }

Notice that while the builder methods return a reference to the created component, they also write the reference directly to the member variable. The returned reference might not be used by the generated code at all (in the constructor or in the builder methods), but you can use it for your purposes. The builder of the main layout is called in the constructor, as explained in Section 7.4.3, “The Constructor”. When you have a layout with nested layout components, the builders of each layout will call the appropriate builder methods of their contained layouts to create their contents.

190

Sub-Component References


Visual User Interface Design with Eclipse (experimental)

7.4.3. The Constructor When you create a new composite component using the wizard, it will create a constructor for the component and fill its basic content. public MyComponent() { buildMainLayout(); setCompositionRoot(mainLayout); // TODO add user code here }

The most important thing to do in the constructor is to set the composition root of the CustomComponent with the setCompositionRoot() (see Section 5.21, â&#x20AC;&#x153;Component Composition with CustomComponentâ&#x20AC;? for more details on the composition root). The generated constructor first builds the root layout of the composite component with buildMainLayout() and then uses the mainLayout reference. The editor will not change the constructor afterwards, so you can safely change it as you want. The editor does not allow changing the member variable holding a reference to the root layout, so it is always named mainLayout.

The Constructor

191


192


Chapter 8

Themes 8.1. Overview ................................................................................................ 193 8.2. Introduction to Cascading Style Sheets ................................................. 195 8.3. Creating and Using Themes .................................................................. 200 8.4. Creating a Theme in Eclipse .................................................................. 204 This chapter provides details about using and creating themes that control the visual look of web applications. Themes consist of Cascading Style Sheets (CSS) and other theme resources such as images. We provide an introduction to CSS, especially concerning the styling of HTML by element classes.

8.1. Overview Vaadin separates the appearance of the user interface from its logic using themes. Themes can include CSS style sheets, custom HTML layouts, and any necessary graphics. Theme resources can also be accessed from an application as ThemeResource objects. Custom themes are placed under the WebContents/VAADIN/themes/ folder of the web application. This location is fixed -- the VAADIN folder specifies that these are static resources specific to Vaadin. The folder should normally contain also the built-in themes, although you can let them be loaded dynamically from the Vaadin JAR (even though that is somewhat inefficient). Figure 8.1, â&#x20AC;&#x153;Contents of a Themeâ&#x20AC;? illustrates the contents of a theme.

Book of Vaadin

193


Themes

Figure 8.1. Contents of a Theme

The name of a theme folder defines the name of the theme. The name is used in the setTheme() call. A theme must contain the styles.css stylesheet, but other contents have free naming. We suggest a convention for naming the folders as img for images, layouts for custom layouts, and css for additional stylesheets. Custom themes that use an existing complete theme need to inherit the theme. See Section 8.3.2, “Built-in Themes” and Section 8.3.4, “Theme Inheritance” for details on inheriting a theme. Copying and modifying a complete theme is also possible, but it may need more work to maintain if the modifications are small. You use a theme with a simple setTheme() method call for the Application object as follows: public class MyApplication extends com.vaadin.Application { public void init() { setTheme("demo"); ... } }

An application can use different themes for different users and switch between themes during execution. For smaller changes, a theme can contain alternate styles for user interface components, which can be changed as needed. In addition to style sheets, a theme can contain HTML templates for custom layouts used with CustomLayout. See Section 6.13, “Custom Layouts” for details. Resources provided in a theme can also be accessed using the ThemeResource class, as described in Section 4.5.4, “Theme Resources”. This allows using theme resources, such as images, for example in Embedded objects and other objects that allow inclusion of images using resources.

194

Overview


Themes

8.2. Introduction to Cascading Style Sheets Cascading Style Sheets or CSS is a technique to separate the appearance of a web page from the content represented in HTML or XHTML. Let us give a short introduction to Cascading Style Sheets and look how they are relevant to software development with Vaadin.

8.2.1. Basic CSS Rules A style sheet is a file that contains a set of rules. Each rule consists of one or more selectors, separated with commas, and a declaration block enclosed in curly braces. A declaration block contains a list of property statements. Each property has a label and a value, separated with a colon. A property statement ends with a semicolon. Let us look at an example: p, td { color: blue; } td { background: yellow; font-weight: bold; }

In the example above, p and td are element type selectors that match with <p> and <td> elements in HTML, respectively. The first rule matches with both elements, while the second matches only with <td> elements. Let us assume that you have saved the above style sheet with the name mystylesheet.css and consider the following HTML file located in the same folder. <html> <head> <link rel="stylesheet" type="text/css" href="mystylesheet.css"/> </head> <body> <p>This is a paragraph</p> <p>This is another paragraph</p> <table> <tr> <td>This is a table cell</td> <td>This is another table cell</td> </tr> </table> </body> </html>

The <link> element defines the style sheet to use. The HTML elements that match the above rules are emphasized. When the page is displayed in the browser, it will look as shown in the figure below.

Figure 8.2. Simple Styling by Element Type

Introduction to Cascading Style Sheets

195


Themes

CSS has an inheritance mechanism where contained elements inherit the properties of their parent elements. For example, let us change the above example and define it instead as follows: table { color: blue; background: yellow; }

All elements contained in the <table> element would have the same properties. For example, the text in the contained <td> elements would be in blue color. Each HTML element type accepts a certain set of properties. The <div> elements are generic elements that can be used to create almost any layout and formatting that can be created with a specific HTML element type. Vaadin uses <div> elements extensively, especially for layouts. Matching elements by their type is, however, rarely if ever used in style sheets for Vaadin components or Google Web Toolkit widgets.

8.2.2. Matching by Element Class Matching HTML elements by the class attribute of the elements is the most relevant form of matching with Vaadin. It is also possible to match with the identifier of a HTML element. The class of an HTML element is defined with the class attribute as follows: <html> <body> <p class="normal">This is the first paragraph</p> <p class="another">This is the second paragraph</p> <table> <tr> <td class="normal">This is a table cell</td> <td class="another">This is another table cell</td> </tr> </table> </body> </html>

The class attributes of HTML elements can be matched in CSS rules with a selector notation where the class name is written after a period following the element name. This gives us full control of matching elements by their type and class. p.normal p.another td.normal td.another

{color: red;} {color: blue;} {background: pink;} {background: yellow;}

The page would look as shown below:

Figure 8.3. Matching HTML Element Type and Class

196

Matching by Element Class


Themes

We can also match solely by the class by using the universal selector * for the element name, for example *.normal. The universal selector can also be left out altogether so that we use just the class name following the period, for example .normal. .normal { color: red; } .another { blackground: yellow; }

In this case, the rule will match with all elements of the same class regardless of the element type. The result is shown in Figure 8.4, â&#x20AC;&#x153;Matching Only HTML Element Classâ&#x20AC;?. This example illustrates a technique to make style sheets compatible regardless of the exact HTML element used in drawing a component.

Figure 8.4. Matching Only HTML Element Class

To assure compatibility, we recommend that you use only matching based on the element classes and do not match for specific HTML element types in CSS rules, because either Vaadin or GWT may use different HTML elements to render some components in the future. For example, IT Mill Toolkit Release 4 used <div> elements extensively for layout components. However, IT Mill Toolkit Release 5 and Vaadin use GWT to render the components, and GWT uses the <table> element to implement most layouts. Similarly, IT Mill Toolkit Release 4 used <div> element also for buttons, but in Release 5, GWT uses the <button> element. Vaadin has little control over how GWT renders its components, so we can not guarantee compatibility in different versions of GWT. However, both <div> and <table> as well as <tr> and <td> elements accept most of the same properties, so matching only the class hierarchy of the elements should be compatible in most cases.

8.2.3. Matching by Descendant Relationship CSS allows matching HTML by their containment relationship. For example, consider the following HTML fragment: <body> <p class="mytext">Here is some text inside a paragraph element</p> <table class="mytable"> <tr> <td class="mytext">Here is text inside a table and inside a td element.</td> </tr> </table> </body>

Matching by the class name .mytext alone would match both the <p> and <td> elements. If we want to match only the table cell, we could use the following selector: .mytable .mytext {color: blue;}

Matching by Descendant Relationship

197


Themes

To match, a class listed in a rule does not have to be an immediate descendant of the previous class, but just a descendant. For example, the selector ".v-panel .v-button" would match all elements with class .v-button somewhere inside an element with class .v-panel. Let us give an example with a real case. Consider the following Vaadin component. public class LoginBox extends CustomComponent { Panel panel = new Panel("Log In"); public LoginBox () { setCompositionRoot(panel); panel.addComponent(new TextField("Username:")); panel.addComponent(new TextField("Password:")); panel.addComponent(new Button("Login")); } }

The component will look by default as shown in the following figure.

Figure 8.5. Themeing Login Box Example with 'runo' theme.

Now, let us look at the HTML structure of the component. The following listing assumes that the application contains only the above component in the main window of the application. <body> <div id="itmtk-ajax-window"> <div> <div class="v-orderedlayout"> <div> <div class="v-panel"> <div class="v-panel-caption">Log In</div> <div class="v-panel-content"> <div class="v-orderedlayout"> <div> <div> <div class="v-caption"> <span>Username:</span> </div> </div> <input type="text" class="v-textfield"/> </div> <div> <div> <div class="v-caption"> <span>Password:</span> </div> </div> <input type="password" class="v-textfield"/> </div> <div>

198

Matching by Descendant Relationship


Themes

<button type="button" class="v-button">Login</button> </div> </div> </div> <div class="v-panel-deco"/> </div> </div> </div> </div> </div> </body>

Now, consider the following theme where we set the backgrounds of various elements. .v-panel .v-panel-caption { background: #80ff80; /* pale green */ } .v-panel .v-panel-content { background: yellow; } .v-panel .v-textfield { background: #e0e0ff; /* pale blue */ } .v-panel .v-button { background: pink; }

The coloring has changed as shown in the following figure.

Figure 8.6. Themeing Login Box Example with Custom Theme

An element can have multiple classes separated with a space. With multiple classes, a CSS rule matches an element if any of the classes match. This feature is used in many Vaadin components to allow matching based on the state of the component. For example, when the mouse is over a Link component, over class is added to the component. Most of such styling is a feature of Google Web Toolkit.

8.2.4. Notes on Compatibility CSS was first proposed in 1994. The specification of CSS is maintained by the CSS Working Group of World Wide Web Consortium (W3C). Its versions are specified as levels that build upon the earlier version. CSS Level 1 was published in 1996, Level 2 in 1998. Development of CSS Level 3 was started in 1998 and is still under way.

Notes on Compatibility

199


Themes

While the support for CSS has been universal in all graphical web browsers since at least 1995, the support has been very incomplete at times and there still exists an unfortunate number of incompatibilities between browsers. While we have tried to take these incompatibilities into account in the built-in themes in Vaadin, you need to consider them while developing custom themes. Compatibility issues are detailed in various CSS handbooks.

8.3. Creating and Using Themes Custom themes are placed in VAADIN/themes folder of the web application (in the WebContent directory) as illustrated in Figure 8.1, “Contents of a Theme”. This location is fixed. You need to have a theme folder for each theme you use in your application, although applications rarely need more than a single theme. For example, if you want to define a theme with the name mytheme, you will place it in folder VAADIN/themes/mytheme. A custom theme must also inherit a built-in theme, as shown in the example below: @import "../reindeer/styles.css"; .v-app { background: yellow; }

Vaadin 6.0 includes two built-in themes: reindeer and runo. The latter is a compatibility theme for IT Mill Toolkit 5; there is no longer a "default" theme. See Section 8.3.2, “Built-in Themes” and Section 8.3.4, “Theme Inheritance” below for details on inheriting themes.

8.3.1. Styling Standard Components Each user interface component in Vaadin has a CSS style class that you can use to control the appearance of the component. Some components have additional sub-elements that also allow styling. Table 8.1, “Default CSS Style Names of Vaadin Components” lists the style classes of all Vaadin components, together with their client-side widgets. Notice that a single server-side component can have multiple client-side implementations. For example, a Button can be rendered on the client side either as a regular button or a check box, depending on the switchMode attribute of the button. For details regarding the mapping to client-side components, see Section 10.5, “Defining a Widget Set”. Each client-side component type has its own style class and a number of additional classes that depend on the client-side state of the component. For example, a text field will have v-textfield-focus class when mouse pointer hovers over the component. This state is purely on the client-side and is not passed to the server.

200

Creating and Using Themes


Themes

Table 8.1. Default CSS Style Names of Vaadin Components Server-Side Component Client-Side Widget

CSS Class Name

AbsoluteLayout

VAbsoluteLayout

v-absolutelayout

Accordion

VAccordion

v-accordion

Button

VButton

v-button

CheckBox

VCheckBox

v-checkbox

CssLayout

VCssLayout

v-csslayout

CustomComponent

VCustomComponent

v-customcomponent

CustomLayout

VCustomLayout

v-customlayout

DateField

VDateField

v-datefield

VCalendar

v-datefield-entrycalendar

VDateFieldCalendar

v-datefield-calendar

VPopupCalendar

v-datefield-calendar

VTextualDate Embedded

VEmbedded

-

Form

VForm

v-form

FormLayout

VFormLayout

-

GridLayout

VGridLayout

-

Label

VLabel

v-label

Link

VLink

v-link

OptionGroup

VOptionGroup

v-select-optiongroup

HorizontalLayout

VHorizontalLayout

v-horizontallayout

VerticalLayout

VVerticalLayout

v-verticallayout

Panel

VPanel

v-panel

VListSelect

v-listselect

VFilterSelect

v-filterselect

Slider

VSlider

v-slider

SplitPanel

VSplitPanel

-

Select

VSplitPanelHorizontal VSplitPanelVertical

-

VScrollTable

v-table

VTablePaging

v-table

TabSheet

VTabSheet

v-tabsheet

TextField

VTextField

v-textfield

Table

VTextArea VPasswordField Tree

VTree

v-tree

TwinColSelect

VTwinColSelect

v-select-twincol

Styling Standard Components

201


Themes

Server-Side Component Client-Side Widget

CSS Class Name

Upload

VUpload

-

Window

VWindow

v-window

-

CalendarEntry

-

-

CalendarPanel

v-datefield-calendarpanel

-

ContextMenu

v-contextmenu

-

VUnknownComponent vaadin-unknown

-

VView

-

-

Menubar

gwt-MenuBar

-

MenuItem

gwt-MenuItem

-

Time

v-datefield-time

Please see the documentation of the particular components for a listing of possible sub-component styles. Some client-side components can be shared by different server-side components. There is also the VUnknownComponent, which is a component that indicates an internal error in a situation where the server asked to render a component which is not available on the client-side.

8.3.2. Built-in Themes Vaadin currently includes two built-in themes: reindeer and runo. The latter is the default theme for IT Mill Toolkit 5 (where its name is "default"); the default theme in Vaadin 6.0 is reindeer. The built-in themes are provided in the respective VAADIN/themes/reindeer/styles.css and VAADIN/themes/runo/styles.css stylesheets in the Vaadin library JAR. These stylesheets are compilations of the separate stylesheets for each component in the corresponding subdirectory. The stylesheets are compiled to a single file for efficiency: the browser needs to load just a single file. Various constants related to the built-in themes are defined in the theme classes in com.vaadin.ui.themes package. These are mostly special style names for specific components. setTheme("runo"); Panel panel = new Panel("Regular Panel in the Runo Theme"); panel.addComponent(new Button("Regular Runo Button")); // A button with the "small" style Button smallButton = new Button("Small Runo Button"); smallButton.addStyleName(Runo.BUTTON_SMALL); Panel lightPanel = new Panel("Light Panel"); lightPanel.addStyleName(Runo.PANEL_LIGHT); lightPanel.addComponent(new Label("With addStyleName(\"light\")"));

The example with the Runo theme is shown in Figure 8.7, â&#x20AC;&#x153;Runo Themeâ&#x20AC;?.

202

Built-in Themes


Themes

Figure 8.7. Runo Theme

Serving Built-In Themes Statically The built-in themes included in the Vaadin library JAR are served dynamically from the JAR by the servlet. Serving themes and widget sets statically by the web server is more efficient.You only need to extract the VAADIN/ directory from the JAR under your WebContent directory. Just make sure to update it if you upgrade to a newer version of Vaadin. Creation of a default theme for custom GWT widgets is described in Section 10.3.3, “Styling GWT Widgets”.

8.3.3. Using Themes Using a theme is simple, you only need to set the theme with setTheme() in the application object. The Eclipse wizard for creating custom Vaadin themes automatically adds such call in the init() method of the application class, as explained in Section 8.4, “Creating a Theme in Eclipse”. Defining the appearance of a user interface component is fairly simple. First, you create a component and add a custom style name for it with addStyleName(). Then you write the CSS element that defines the formatting for the component.

8.3.4. Theme Inheritance When you define your own theme, you will need to inherit a built-in theme (unless you just copy the built-in theme). Inheritance in CSS is done with the @import statement. In the typical case, when you define your own theme, you inherit a built-in theme as follows: @import "../reindeer/styles.css"; .v-app { background: yellow; }

You can even create a deep hierarchy of themes by inheritance. Such a solution is often useful if you have some overall theme for your application and a slightly modified theme for different user classes. You can even make it possible for each user to have his or her own theme. For example, let us assume that we have the base theme of an application with the name myapp and a specific myapp-student theme for users with the student role. The stylesheet of the base

Using Themes

203


Themes

theme would be located in themes/myapp/styles.css. We can then "inherit" it in themes/myapp-student/styles.css with a simple @import statement: @import "../myapp/styles.css"; .v-app { background: green; }

This would make the page look just as with the base theme, except for the green background. You could use the theme inheritance as follows: public class MyApplication extends com.vaadin.Application { public void init() { setTheme("myapp"); ... } public void login(User user) { if (user.role == User.ROLE_STUDENT) setTheme("myapp-student"); ... } public void logout() { setTheme("myapp"); ... } }

In the above example, the User class is an imaginary class assumed to hold user information.

8.4. Creating a Theme in Eclipse The Eclipse plugin provides a wizard for creating custom themes. Do the following steps to create a new theme. 1. Select File → New → Other... in the main menu or right-click the Project Explorer and select New → Other.... A window will open. 2. In the Select a wizard step, select the Vaadin → Vaadin Theme wizard.

Click Next to proceed to the next step. 3. In the Create a new Vaadin theme step, you have the following settings: Project (mandatory)

204

The project in which the theme should be created.

Creating a Theme in Eclipse


Themes

Theme name (mandatory)

The theme name is used as the name of the theme folder and in a CSS tag (prefixed with "v-theme-"), so it must be a proper identifier. Only latin alphanumerics, underscore, and minus sign are allowed.

Modify application classes to use theme (optional)

The setting allows the wizard to write a code statement that enables the theme in the constructor of the selected application class(es). If you need to control the theme with dynamic logic, you can leave the setting unchecked or change the generated line later.

Click Finish to create the theme. The wizard creates the theme folder under the WebContent/VAADIN/themes folder and the actual style sheet as styles.css, as illustrated in Figure 8.8, “Newly Created Theme”.

Figure 8.8. Newly Created Theme

The created theme inherits a built-in base theme with an @import statement. See the explanation of theme inheritance in Section 8.3, “Creating and Using Themes”. Notice that the reindeer theme is not located in the widgetsets folder, but in the Vaadin JAR. See Section 8.3.2, “Builtin Themes” for information for serving the built-in themes. If you selected an application class or classes in the Modify application classes to use theme in the theme wizard, the wizard will add the following line at the end of the init() method of the application class(es): setTheme("myprojecttheme");

Notice that renaming a theme by changing the name of the folder will not change the setTheme() calls in the application classes or vise versa. You need to change such references to theme names in the calls manually.

Creating a Theme in Eclipse

205


206


Chapter 9

Binding Components to Data 9.1. Overview ................................................................................................ 207 9.2. Properties .............................................................................................. 209 9.3. Holding properties in Items .................................................................... 212 9.4. Collecting items in Containers ............................................................... 214 This chapter describes the Vaadin Data Model and shows how you can use it to bind components directly to data sources, such as database queries.

9.1. Overview The Vaadin Data Model is one of the core concepts of the library. To allow the view (user interface components) to access the data model of an application directly, we have introduced a standard data interface. The model allows binding user interface components directly to the data that they display and possibly allow to edit. There are three nested levels of hierarchy in the data model: property, item, and container. Using a spreadsheet application as an analogy, these would correspond to a cell, a row, and a table, respectively.

Book of Vaadin

207


Binding Components to Data

Figure 9.1. Vaadin Data Model

The Data Model is realized as a set of interfaces in the com.vaadin.data package. The package contains the Property, Item, and Container interfaces, along with a number of more specialized interfaces and classes. Notice that the Data Model does not define data representation, but only interfaces. This leaves the representation fully to the implementation of the containers. The representation can be almost anything, such as a plain old Java object (POJO) structure, a filesystem, or a database query. The Data Model is used heavily in the core user interface components of Vaadin, especially the field components, that is, components that implement the Field interface or more typically extend AbstractField, which defines many common features. A key feature of all the built-in field components is that they can either maintain their data by themselves or be bound to an external data source. The value of a field is always available through the Property interface. As more than one component can be bound to the same data source, it is easy to implement various viewereditor patterns. The relationships of the various interfaces are shown in Figure 9.2, â&#x20AC;&#x153;Interface Relationships in Vaadin Data Modelâ&#x20AC;?; the value change event and listener interfaces are shown only for the Property interface, while the notifier interfaces are omitted altogether.

208

Overview


Binding Components to Data

Figure 9.2. Interface Relationships in Vaadin Data Model

The Data Model has many important and useful features, such as support for change notification. Especially containers have many helper interfaces, including ones that allow indexing, ordering, sorting, and filtering the data. Also Field components provide a number of features involving the data model, such as buffering, validation, and lazy loading. In addition to the interfaces, Vaadin library provides a number of built-in implementations of the interfaces. The built-in implementations are used as the default data models in many field components.

9.2. Properties The Property interface is the base of the Vaadin Data Model. It provides a standardized API for a single data object that can be read (get) and written (set). A property is always typed, but can optionally support data type conversions. The type of a property can be any Java class. Optionally, properties can provide value change events for following their changes. The value of a property is written with setValue() and read with getValue(). The return value is a generic Object reference, so you need to cast it to the proper type. The type can be acquired with getType(). final TextField tf = new TextField("Name"); // Set the value tf.setValue("The text field value"); // When the field value is edited by the user tf.addListener(new Property.ValueChangeListener() { public void valueChange(ValueChangeEvent event) { // Get the value and cast it to proper type String value = (String) tf.getValue(); // Do something with it

Properties

209


Binding Components to Data

layout.addComponent(new Label(value)); } });

Changes in the property value usually emit a ValueChangeEvent, which can be handled with a ValueChangeListener.The event object provides reference to the property with getProperty(). Properties are in themselves unnamed.They are collected in items, which associate the properties with names: the Property Identifiers or PIDs. Items can be further contained in containers and are identified with Item Identifiers or IIDs. In the spreadsheet analogy, Property Identifiers would correspond to column names and Item Identifiers to row names. The identifiers can be arbitrary objects, but must implement the equals(Object) and hashCode() methods so that they can be used in any standard Java Collection. The Property interface can be utilized either by implementing the interface or by using some of the built-in property implementations. Vaadin includes a Property interface implementation for arbitrary function pairs and bean properties, with the MethodProperty class, and for simple object properties, with the ObjectProperty class, as described later. In addition to the simple components, many selection components such as Select, Table, and Tree provide their current selection through the Property property. In single selection mode, the property is a single item identifier, while in multiple selection mode it is a set of item identifiers. Please see the documentation of the selection components for further details. Components that can be bound to a property have an internal default data source object, typically a ObjectProperty, which is described later. As all such components are viewers or editors, also described later, so you can rebind a component to any data source with setPropertyDataSource().

9.2.1. Property Viewers and Editors The most important function of the Property as well as of the other data model interfaces is to connect classes implementing the interface directly to editor and viewer classes. This means connecting a data source (model) to a user interface component (views) to allow editing or viewing the data model. A property can be bound to a component implementing the Viewer interface with setPropertyDataSource(). // Have a data model ObjectProperty property = new ObjectProperty("Hello", String.class); // Have a component that implements Viewer Label viewer = new Label(); // Bind it to the data viewer.setPropertyDataSource(property);

You can use the same method in the Editor interface to bind a component that allows editing a particular property type to a property. // Have a data model ObjectProperty property = new ObjectProperty("Hello", String.class); // Have a component that implements Viewer TextField editor = new TextField("Edit Greeting");

210

Property Viewers and Editors


Binding Components to Data

// Bind it to the data editor.setPropertyDataSource(property);

As all field components implement the Property interface, you can bind any component implementing the Viewer interface to any field, assuming that the viewer is able the view the object type of the field. Continuing from the above example, we can bind a Label to the TextField value: Label viewer = new Label(); viewer.setPropertyDataSource(editor); // The value shown in the viewer is updated immediately // after editing the value in the editor (once it // loses the focus) editor.setImmediate(true);

9.2.2. ObjectProperty Implementation The ObjectProperty class is a simple implementation of the Property interface that allows storing an arbitrary Java object. // Have a component that implements Viewer interface final TextField tf = new TextField("Name"); // Have a data model with some data String myObject = "Hello"; // Wrap it in an ObjectProperty ObjectProperty property = new ObjectProperty(myObject, String.class); // Bind the property to the component tf.setPropertyDataSource(property);

9.2.3. Implementing the Property Interface Implementation of the Property interface requires defining setters and getters for the value and the read-only mode. Only a getter is needed for the property type, as the type is often fixed in property implementations. The following example shows a simple implementation of the Property interface: class MyProperty implements Property { Integer data = 0; boolean readOnly = false; // Return the data type of the model public Class<?> getType() { return Integer.class; } public Object getValue() { return data; } // Override the default implementation in Object @Override public String toString() { return Integer.toHexString(data); } public boolean isReadOnly() { return readOnly; }

ObjectProperty Implementation

211


Binding Components to Data

public void setReadOnly(boolean newStatus) { readOnly = newStatus; } public void setValue(Object newValue) throws ReadOnlyException, ConversionException { if (readOnly) throw new ReadOnlyException(); // Already the same type as the internal representation if (newValue instanceof Integer) data = (Integer) newValue; // Conversion from a string is required else if (newValue instanceof String) try { data = Integer.parseInt((String) newValue, 16); } catch (NumberFormatException e) { throw new ConversionException(); } else // Don't know how to convert any other types throw new ConversionException(); // Reverse decode the hexadecimal value } } // Instantiate the property and set its data MyProperty property = new MyProperty(); property.setValue(42); // Bind it to a component final TextField tf = new TextField("Name", property);

The components get the displayed value by the toString() method, so it is necessary to override it. To allow editing the value, value returned in the toString() must be in a format that is accepted by the setValue() method, unless the property is read-only. The toString() can perform any type conversion necessary to make the internal type a string, and the setValue() must be able to make a reverse conversion. The implementation example does not notify about changes in the property value or in the readonly mode. You should normally also implement at least the Property.ValueChangeNotifier and Property.ReadOnlyStatusChangeNotifier. See the ObjectProperty class for an example of the implementation.

9.3. Holding properties in Items The Item interface provides access to a set of named properties. Each property is identified by a property identifier (PID) and a reference to such a property can be queried from an Item with getItemProperty() using the identifier. Examples on the use of items include rows in a Table, with the properties corresponding to table columns, nodes in a Tree, and the the data bound to a Form, with item's properties bound to individual form fields. Items are generally equivalent to objects in the object-oriented model, but with the exception that they are configurable and provide an event handling mechanism. The simplest way to utilize Item interface is to use existing implementations. Provided utility classes include a configurable property set (PropertysetItem) and a bean-to-item adapter (BeanItem). Also, a Form implements the interface and can therefore be used directly as an item.

212

Holding properties in Items


Binding Components to Data

In addition to being used indirectly by many user interface components, items provide the basic data model underlying the Form component. In simple cases, forms can even be generated automatically from items. The properties of the item correspond to the fields of the form. The Item interface defines inner interfaces for maintaining the item property set and listening changes made to it. PropertySetChangeEvent events can be emitted by a class implementing the PropertySetChangeNotifier interface. They can be received through the PropertySetChangeListener interface.

9.3.1. The PropertysetItem Implementation The PropertysetItem is a generic implementation of the Item interface that allows storing properties. The properties are added with addItemProperty(), which takes a name and the property as parameters. The following example demonstrates a typical case of collecting ObjectProperty properties in an item: PropertysetItem item = new PropertysetItem(); item.addItemProperty("name", new ObjectProperty("Zaphod")); item.addItemProperty("age", new ObjectProperty(42)); // Bind it to a component Form form = new Form(); form.setItemDataSource(item);

9.3.2. Wrapping a Bean in a BeanItem The BeanItem implementation of the Item interface is a wrapper for Java Bean objects. In fact, only the setters and getters are required while serialization and other bean features are not, so you can wrap almost any POJOs with minimal requirements. // Here is a bean (or more exactly a POJO) class Person { String name; int age; public String getName() { return name; } public void setName(String name) { this.name = name; } public Integer getAge() { return age; } public void setAge(Integer age) { this.age = age.intValue(); } } // Create an instance of the bean Person bean = new Person(); // Wrap it in a BeanItem BeanItem<Person> item = new BeanItem<Person>(bean); // Bind it to a component

The PropertysetItem Implementation

213


Binding Components to Data

Form form = new Form(); form.setItemDataSource(item);

You can use the getBean() method to get a reference to the underlying bean.

9.4. Collecting items in Containers Container is the highest-level of the data model interfaces supported by Vaadin. It provides a very flexible way of managing a set of items that share common properties. Each item is identified by an item identifier or IID. Items can be added to a container with the addItem() method. Notice that the actual item is not passed as a parameter to the method, only the item ID, as the interface assumes that the implementation knows how to create the item. The parameterless version of the method uses an automatically generated item ID. Implementations can provide methods to add externally created items, or they can assume that the item ID is also the item itself. Properties can be requested from container by first requesting an item with getItem() and then getting the properties from the item with getItemProperty(). You can also get a property directly by the item and property ids with getContainerProperty(). The Container interface was designed with flexibility and efficiency in mind. It contains inner interfaces that containers can optionally implement for ordering the items sequentially, indexing the items, and accessing them hierarchically. Such ordering models provide the basis for the Table, Tree, and Select components. As with other data model interfaces, the Container supports events for notifying about changes made to their contents. As containers can be unordered, ordered, indexed, or hierarchical, they can interface practically any kind of data representation. Vaadin includes data connectors for some common data sources, such as the simple tabular data, with IndexedContainer, and the filesystem, with FilesystemContainer. In addition to generic container implementations, also many user interface components are containers as themselves, in addition to being properties. This is especially true for selection components, that is, those that implement Select, because they are containers that contain selectable items. Their property is the currently selected item. This is useful as it allows binding components to view and updating each others' data directly, and makes it easy to reuse already constructed data models, for example, a form could edit a row (item) of a table directly, and the table could use a database container as its underlying container. The fields of the form would correspond to the properties of the item, that is, the cells of the table row. The library contains a set of utilities for converting between different container implementations by adding external ordering or hierarchy into existing containers. In-memory containers implementing indexed and hierarchical models provide easy-to-use tools for setting up in-memory data storages. Such default container implementations include IndexedContainer, which can be thought of as a generalization of a two-dimensional data table, and BeanItemContainer, which maps standard Java objects (beans) to items of an indexed container. In addition, the built-in containers include a hierarchical container for direct file system browsing.

9.4.1. Iterating Over a Container As the items in a Container are not necessarily indexed, iterating over the items has to be done using an Iterator. The getItemIds() method of Container returns a Collection of item identifiers over which you can iterate. The following example demonstrates a typical case where you

214

Collecting items in Containers


Binding Components to Data

iterate over the values of check boxes in a column of a Table component. The context of the example is the example used in Section 5.12, â&#x20AC;&#x153;Tableâ&#x20AC;?. // Collect the results of the iteration into this string. String items = ""; // Iterate over the item identifiers of the table. for (Iterator i = table.getItemIds().iterator(); i.hasNext();) { // Get the current item identifier, which is an integer. int iid = (Integer) i.next(); // Now get the actual item from the table. Item item = table.getItem(iid); // And now we can get to the actual checkbox object. Button button = (Button) (item.getItemProperty("ismember").getValue()); // If the checkbox is selected. if ((Boolean)button.getValue() == true) { // Do something with the selected item; collect the // first names in a string. items += item.getItemProperty("First Name") .getValue() + " "; } } // Do something with the results; display the selected items. layout.addComponent (new Label("Selected items: " + items));

Notice that the getItemIds() returns an unmodifiable collection, so the Container may not be modified during iteration. You can not, for example, remove items from the Container during iteration. The modification includes modification in another thread. If the Container is modified during iteration, a ConcurrentModificationException is thrown and the iterator may be left in an undefined state.

Iterating Over a Container

215


216


Chapter 10

Developing Custom Components 10.1. Overview .............................................................................................. 218 10.2. Doing It the Simple Way in Eclipse ...................................................... 220 10.3. Google Web Toolkit Widgets ................................................................ 224 10.4. Integrating a GWT Widget ................................................................... 229 10.5. Defining a Widget Set .......................................................................... 234 10.6. Server-Side Components .................................................................... 235 10.7. Using a Custom Component ............................................................... 237 10.8. GWT Widget Development .................................................................. 239 This chapter describes how you can create custom client-side components as Google Web Toolkit (GWT) widgets and how you integrate them with Vaadin. The client-side implementations of all standard user interface components in Vaadin use the same client-side interfaces and patterns. Google Web Toolkit is intended for developing browser-based user interfaces using the Java language, which is compiled into JavaScript that is executed in the browser. Knowledge of such client-side technologies is usually not needed with Vaadin, as its built-in repertoire of user interface components should be sufficient for most applications. The easiest way to create custom components in Vaadin is to make composite components with the CustomComponent class, as

Book of Vaadin

217


Developing Custom Components

described in Section 5.21, “Component Composition with CustomComponent”. In some cases, however, you may need to make modifications to existing components, integrate existing GWT widgets with your application, or create entirely new ones. Creation of new widgets involves a number of rather intricate tasks. The Vaadin Plugin for Eclipse makes many of the tasks much easier, so if you are using Eclipse and the plugin, you should find Section 10.2, “Doing It the Simple Way in Eclipse” helpful. If you need more background on the architecture, Section 3.4, “Client-Side Engine” gives an introduction to the architecture of the Vaadin Client-Side Engine. If you are new to Google Web Toolkit, Section 3.2.2, “Google Web Toolkit” gives an introduction to GWT and its role in the architecture of Vaadin. On Terminology Google Web Toolkit uses the term widget for user interface components. In this book, we use the term widget to refer to client-side components made with Google Web Toolkit, while using the term component in a general sense and also in the special sense for server-side components.

10.1. Overview The Client-Side Engine of Vaadin runs in the web browser as a JavaScript program and renders the user interface components according to state data received from the server. For each serverside component, there is a client-side widget, which renders the content of the particular component type. The client-side engine and all the built-in client-side widgets of Vaadin have been programmed in Java with GWT, and compiled into JavaScript with the GWT Compiler. Developing custom Vaadin components and integrating existing GWT widgets is easy, requiring only Java programming. You can start with any existing GWT widget or design a new one. To integrate it with Vaadin, you have to implement the Paintable interface of the client-side engine that provides the AJAX communications with the server-side application.You can find the interface in the com.vaadin.terminal.gwt.client package. You can also choose to extend an existing Vaadin client-side widget in the com.vaadin.terminal.gwt.client.ui package. You can find the source code for the built-in widgets in the Vaadin JAR. To use custom widgets, you need to define a widget set that inherits the DefaultWidgetSet, which contains the standard widgets, or some other widget set. You can also define stylesheets for custom widgets. A widget set is defined in a GWT Module Descriptor. For the server-side API, you need a server-side component that can serialize and deserialize its attributes to and from the client-side counterpart. A server-side component usually inherits the AbstractComponent or AbstractField class and implements either the paintContent() or the more generic paint() method to serialize its data to the client. These methods "paint" the component in the browser by generating a UIDL element that is sent to the client. The UIDL element contains all the relevant information about the component, and you can easily add your own attributes to it. Upon reception of UIDL messages, the client-side engine creates or updates user interface widgets as needed. To summarize, you need to do the following: • Implement the Paintable interface of Vaadin in a GWT widget

218

Overview


Developing Custom Components

• Define a widget set that extends an existing widget set with the new widget • Create a default CSS style sheet for the widget set (optional) • Create a GWT Module Descriptor (.gwt.xml) that defines the widget set and the optional style sheet • Create a server-side counterpart for the client-side widget Figure 10.1, “Color Picker Package” illustrates the folder hierarchy of the Color Picker example used in this chapter. The example is available in the demo application of Vaadin with URL /colorpicker/. You can find the full source code of the application in the source folder of the demos in the installation package.

Figure 10.1. Color Picker Package

The ColorPickerApplication.java application provides an example of using the custom ColorPicker component. The source code for the server-side implementation of the component is located in the same folder. The GWT Compiler takes the root folder of the client-side source code as its argument and compiles all the Java source files into JavaScript. A client-side widget set must therefore be contained within a single package, which in the Color Picker example is the com.vaadin.demo.colorpicker.gwt.client package. The inherited widget set and an optional style sheet are specified in a .gwt.xml descriptor for the GWT Compiler. In the example, the client-side widget is split in two classes: GwtColorPicker, a pure GWT widget, and VColorPicker that provides the integration with Vaadin. The default style sheet for the widget set is defined in the descriptor and located in the gwt/public/colorpicker/styles.css subfolder.

Overview

219


Developing Custom Components

10.2. Doing It the Simple Way in Eclipse While you can develop new widgets with any IDE or even without, you may find Eclipse and the Vaadin Plugin for it useful, as it automates all the basic routines of widget development, most importantly the creation of new widgets.

10.2.1. Creating a Widget You can create a new widget as follows: 1. Select File → New → Other... in the main menu or right-click the Project Explorer and select New → Other... or press Ctrl-N to open the New dialog. 2. In the first, Select a wizard step, select Vaadin → Vaadin Widget and click Next.

3. In the New Component wizard step, fill out the target folder, package, and class information.

220

Doing It the Simple Way in Eclipse


Developing Custom Components

Source folder

The root folder of the entire source tree. The default value is the default source tree of your project, and you should normally leave it unchanged unless you have a different project structure.

Package

The parent package under which the new server-side component should be created. If it does not already exist, the .gwt.xml descriptor that defines the widget set will be created under the widgetset subpackage under this package, and the actual new widget under the widgetset.client.ui subpackage.

Name

The class name of the new server-side component. The name of the client-side widget stub (if you have its creation enabled) will be the same but with "V-" prefix, for example, VMycomponent. You can rename the classes afterwards.

Superclass

The superclass of the server-side component. It is AbstractComponent by default, but com.vaadin.ui.AbstractField or com.vaadin.ui.AbstractSelect are other commonly used superclasses. If you are extending an existing component, you should select it as the superclass.You can easily change the superclass later.

Build client-side stub

When this option is selected (strongly recommended), the wizard will build a stub for the client-side widget.

Finally, click Finish to create the new component. The wizard will: â&#x20AC;˘ Create a server-side component stub in the base package â&#x20AC;˘ Create a GWT module descriptor file (.gwt.xml) in the widgetset package under the base package

Creating a Widget

221


Developing Custom Components

• Create a client-side widget stub in the widgetset.client.ui package under the base package • Modify the web.xml deployment descriptor to specify the widget set class name parameter for the application. The structure of the server-side component and the client-side widget, and the serialization of component state between them, is explained in the subsequent sections of this chapter. To compile the widget set, click the Compile widget set button in the Eclipse toolbar. See Section 10.2.2, “Recompiling the Widget Set” for details. After the compilation finishes, you should be able to run your application as before, but using the new widget set. The compilation result is written under the WebContent/VAADIN/widgetsets folder. When you need to recompile the widget set in Eclipse, see Section 10.2.2, “Recompiling the Widget Set”. For detailed information on compiling widget sets, see Section 10.8.4, “Compiling GWT Widget Sets”. The following setting is inserted in the web.xml deployment descriptor to enable the widget set: <init-param> <description>Application widgetset</description> <param-name>widgetset</param-name> <param-value>com.example.myproject.widgetset.MyprojectApplicationWidgetset</param-value> </init-param>

You can refactor the package structure if you find need for it, but GWT compiler requires that the client-side code must always be stored under a package named "client".

10.2.2. Recompiling the Widget Set After you edit a widget, you need to recompile the widget set. Vaadin automatically suggests to compile the widget set every time you save a client-side source file. If this gets annoying, you can disable the automatic recompilation from the Vaadin category in project settings, by selecting the Suspend automatic widgetset builds option. You can recompile the widget set manually by clicking the Compile widgetset button in the Eclipse toolbar, shown in Figure 10.2, “The Compile Widgetset Button in Eclipse Toolbar” while the project is open and selected.

Figure 10.2. The Compile Widgetset Button in Eclipse Toolbar

The compilation progress is shown in the Console panel in Eclipse, as illustrated in Figure 10.3, “Recompiling a Widget Set”. The compilation output is written under the WebContent/VAADIN/widgetsets folder, in a widget set specific folder. You can speed up the compilation significantly by compiling the widget set only for your browser during development. The generated .gwt.xml descriptor stub includes a disabled element that specifies the target browser. See Section 10.5, “Defining a Widget Set” for more details on setting the user-agent property.

222

Recompiling the Widget Set


Developing Custom Components

Figure 10.3. Recompiling a Widget Set

For detailed information on compiling widget sets, see Section 10.8.4, “Compiling GWT Widget Sets”. Should you compile a widget set outside Eclipse, you need to refresh the project by selecting it in Project Explorer and pressing F5.

10.2.3. Plugin Related Project Settings When you have the Eclipse Plugin installed, the project settings will have a Vaadin section, where you can select the Vaadin version and make settings to widget set building. The settings window is shown in Figure 10.4, “Plugin Related Project Settings”.

Figure 10.4. Plugin Related Project Settings

Suspend automatic widgetset builds

Normally, when this option is unselected, Eclipse will suggest to rebuild the widget set automatically every time you save a widget source file. If this gets annoying, you can suspend the automatic building by enabling

Plugin Related Project Settings

223


Developing Custom Components

this option. You then have to click the widget set build button in the Eclipse toolbar. JavaScript style

Normally, GWT outputs obfuscated JavaScript to make the code less readable. The main purpose is to protect the intellectual property, but the obfuscated code is also more compact, reducing the time required to load and parse the files. The Obfuscated mode is the default. The other output types are Pretty, which makes the JavaScript more readable to a human, and Detailed, which is more detailed than the pretty option with, for example, more descriptive variable names.

Compiler threads

You can set the GWT Compiler to use a specific number of threads to use the available processor cores.

Create development mode launch

Clicking this button generates a launch configuration for starting the application in GWT Development Mode. You can use the launch configuration to debug clientside code. See Section 10.8.6, “GWT Development Mode” for detailed information on the GWT Develpment Mode launch configuration.

The plugin will automatically download the GWT version compatible with the selected Vaadin package.

10.3. Google Web Toolkit Widgets Let us take a look into how you create custom GWT widgets. The authoritative sources for developing with GWT are the Google Web Toolkit Developer Guide and Google Web Toolkit Class Reference, both available from the GWT website. Google Web Toolkit offers a variety of ways for creating custom widgets. The easiest way is to create composite widgets by grouping existing basic widgets and adding some interaction logic to them.You can also develop widgets using the lower-level Java interfaces used by the standard GWT widgets or the really low-level JavaScript interfaces. A custom GWT widget needs to find its place in the GWT class hierarchy. Figure 10.5, “GWT Widget Base Class Hierarchy” illustrates the abstract base classes for GWT widgets.

224

Google Web Toolkit Widgets


Developing Custom Components

Figure 10.5. GWT Widget Base Class Hierarchy

Each of the base classes offers various services for different types of widgets. Many custom widgets, such as the Color Picker example below, extend the Composite class to compose the widget from existing GWT widgets.The other base classes offer various features useful for different kinds of widgets. You can also choose to extend an existing GWT widget, as we have done for most of the standard user interface components of Vaadin, or to extend a Vaadin widget.

10.3.1. Extending a Vaadin Widget Extending an existing Vaadin widget is an easy way to add features, such as advanced clientside validation, to existing standard components. Just extend both the server-side component and the client-side widget, and add the needed properties and other functionality. A few guidelines apply for extending existing components: • Do not redefine the client or id member variables in the subclass. They should always be defined as protected and set by the base class where they are defined in updateFromUIDL(). • Call super() as the first thing in the constructor. • In updateFromUIDL(): • Do not call client.updateComponent() as it is already called in the superclass. If the superclass calls it in an undesired way, you have to reimplement the entire functionality and not call super.updateFromUIDL() at all. • Call super.updateFromUIDL(uidl, client), unless you specifically want to change the behaviour of the superclass implementation of the method. In such case, you have to set the client and id yourself. • Be careful about overloading superclass handlers, etc.

Extending a Vaadin Widget

225


Developing Custom Components

10.3.2. Example: A Color Picker GWT Widget In the following example, we present a custom widget composited from the HorizontalPanel, Grid, Button, and Label GWT widgets. This widget does not include any Vaadin integration with the server-side code, which will be added in a separate integration widget later in this chapter. package com.vaadin.demo.colorpicker.gwt.client.ui; import com.google.gwt.event.dom.client.*; import com.google.gwt.user.client.*; import com.google.gwt.user.client.ui.*; /** * A regular GWT component without integration with Vaadin. */ public class GwtColorPicker extends Composite implements ClickHandler { // The currently selected color name to give client-side // feedback to the user. protected Label currentcolor = new Label(); public GwtColorPicker() { // Create a 4x4 grid of buttons with names for 16 colors final Grid grid = new Grid(4, 4); final String[] colors = new String[] { "aqua", "black", "blue", "fuchsia", "gray", "green", "lime", "maroon", "navy", "olive", "purple", "red", "silver", "teal", "white", "yellow" }; int colornum = 0; for (int i = 0; i < 4; i++) { for (int j = 0; j < 4; j++, colornum++) { // Create a button for each color Button button = new Button(colors[colornum]); button.addClickHandler(this); // Put the button in the Grid layout grid.setWidget(i, j, button); // Set the button background colors. DOM.setStyleAttribute(button.getElement(), "background", colors[colornum]); // For dark colors, the button label must be // in white. if ("black navy maroon blue purple" .indexOf(colors[colornum]) != -1) { DOM.setStyleAttribute(button.getElement(), "color", "white"); } } } // Create a panel with the color grid and currently // selected color indicator. final HorizontalPanel panel = new HorizontalPanel(); panel.add(grid); panel.add(currentcolor); // // // // // //

226

Set the class of the color selection feedback box to allow CSS styling. We need to obtain the DOM element for the current color label. This assumes that the <td> element of the HorizontalPanel is the parent of the label element. Notice that the element has no parent before the widget has been

Example: A Color Picker GWT Widget


Developing Custom Components

// added to the horizontal panel. final Element panelcell = DOM.getParent(currentcolor.getElement()); DOM.setElementProperty(panelcell, "className", "colorpicker-currentcolorbox"); // Set initial color. This will be overridden with the // value read from server. setColor("white"); // Composite GWT widgets must call initWidget(). initWidget(panel); } /** Handles click on a color button. */ @Override public void onClick(ClickEvent event) { // Use the button label as the color name to set setColor(((Button) event.getSource()).getText()); } /** Sets the currently selected color. */ public void setColor(String newcolor) { // Give client-side feedback by changing the color // name in the label. currentcolor.setText(newcolor); // Obtain the DOM elements. This assumes that the <td> // element of the HorizontalPanel is the parent of the // caption element. final Element caption = currentcolor.getElement(); final Element cell = DOM.getParent(caption); // Give feedback by changing the background color DOM.setStyleAttribute(cell, "background", newcolor); DOM.setStyleAttribute(caption, "background", newcolor); if ("black navy maroon blue purple" .indexOf(newcolor) != -1) DOM.setStyleAttribute(caption, "color", "white"); else DOM.setStyleAttribute(caption, "color", "black"); } }

You can find the source code in the source folder of the Vaadin installation package, under the Java package com.vaadin.demo.colorpicker. This example demonstrates one reason for making a custom widget: it provides client-side feedback to the user in a way that would not be possible or at least practical from server-side code. Server-side code can only select a static CSS style or a theme, while on the client-side we can manipulate styles of HTML elements flexibly. Notice that manipulation of the DOM tree depends somewhat on the browser. In this example, the manipulation should be rather compatible, but in some cases there could be problems. Thd standard GWT and Vaadin widgets handle many of such compatibility issues, but when doing low-level operations such as DOM manipulation, you may need to consider browser compatibility. The structure of the DOM tree depends on how GWT renders its widgets in a specific browser. It is also not guaranteed that the rendering does not change in future releases of GWT. You should therefore make as few assumptions regarding the DOM structure as possible. Unfortunately, GWT does not provide a way to set the style of, for example, cells of layout elements. The above example therefore assumes that the Grid is a table and the <button> elements are inside <td> elements of the table. See Section 10.3.3, â&#x20AC;&#x153;Styling GWT Widgetsâ&#x20AC;? below for more details on compatibility.

Example: A Color Picker GWT Widget

227


Developing Custom Components

The widget will look as shown in Figure 10.6, “Color Picker Widget Without Styling”.

Figure 10.6. Color Picker Widget Without Styling

As you may notice, the widget will look rather uninviting without CSS styling. We will next look how to define a default style for a GWT widget.

10.3.3. Styling GWT Widgets GWT renders its widgets in the DOM tree of the web browser as HTML elements. Therefore, you can define their style with Cascading Style Sheets (CSS), just as in HTML. The GWT Compiler supports packaging style sheets from the source package tree. The style sheet is defined in the .gwt.xml GWT module descriptor file (see Section 10.5, “Defining a Widget Set” for details). <!-- Default theme for the widget set. --> <stylesheet src="colorpicker/styles.css"/>

The style sheet path is relative to the public folder under the folder containing the .gwt.xml file. In Eclipse, you can add the folders as regular folders instead of Java packages. Let us define the colorpicker/styles.css as follows. /* Set style for the color picker table. * This assumes that the Grid layout is rendered * as a HTML <table>. */ table.example-colorpicker { border-collapse: collapse; border: 0px; } /* Set color picker button style. * This does not make assumptions about the HTML * element tree as it only uses the class attributes * of the elements. */ .example-colorpicker .gwt-Button { height: 60px; width: 60px; border: none; padding: 0px; } /* Set style for the right-hand box that shows the * currently selected color. While this may work for * other implementations of the HorizontalPanel as well, * it somewhat assumes that the layout is rendered * as a table where cells are <td> elements. */ .colorpicker-currentcolorbox { width: 240px; text-align: center; /* Must be !important to override GWT styling: */ vertical-align: middle !important; }

228

Styling GWT Widgets


Developing Custom Components

The stylesheet above makes some assumptions regarding the HTML element structure. First, it assumes that the Grid layout is a table. Second, the custom class name, colorpicker-currentcolorbox, of the right-hand HorizontalPanel cell was inserted in the DOM representation of the widget in the GwtColorPicker implementation. Styling a button makes less assumptions. Using only class names instead of specific element names may make a stylesheet more compatible if the HTML representation is different in different browsers or changes in the future.

Figure 10.7. Color Picker Widget With Styling

10.4. Integrating a GWT Widget Integration of GWT widgets with Vaadin can be done in two basic ways: by modifying the original widget or by extending it and adding the integration code in the subclass. The latter way is actually the way the standard client-side components in Vaadin are done: they simply inherit the corresponding standard GWT widgets. For example, VButton inherits GWT Button. The client-side integration code has the following tasks: • Receive component state from server • Send state changes caused by user interaction to server • Manage CSS style class The integration is broken down in the following sections into server-client deserialization done in updateFromUIDL() and client-server serialization done with updateVariable().The complete example of the integration of the Color Picker widget is given at the end of this section. If you are using the Eclipse IDE, the Vaadin Plugin for Eclipse allows easy creation of a stub for a new widget, alongside its server-side component. It also manages the widget set for you automatically. See Section 10.2.1, “Creating a Widget” for detailed instructions. Naming Conventions While the use of Vaadin does not require the use of any particular naming conventions for GWT widgets, some notes regarding naming may be necessary. Even though Java package names make it possible to use identical class names in the same context, it may be useful to try to make them more distinctive to avoid any inconveni-

Integrating a GWT Widget

229


Developing Custom Components

ence. GWT uses plain names for its standard widgets, such as Button. The standard components of Vaadin use identical or similar names, but that does not cause any inconvenience, because the GWT widgets and server-side components of Vaadin are never used in the same context. For the client-side components of Vaadin, we use the "V" prefix, for example VButton. In the Color Picker example, we use GwtColorPicker for the GWT widget, VColorPicker for the integration implementation, and ColorPicker for the server-side component. You may wish to follow similar conventions. Notice that the naming convention changed when IT Mill Toolkit was renamed as Vaadin. The prefix for client-side widgets in IT Mill Toolkit was I, which was changed to V in Vaadin. Similarly, CSS style name prefixes were changed from i- to v-.

10.4.1. Deserialization of Component State from Server To receive data from the server, a widget must implement the Paintable interface and its updateFromUIDL() method. The idea is that the method "paints" the user interface description by manipulating the HTML tree on the browser.Typically, when using composite GWT components, most of the DOM tree manipulation is done by standard GWT widgets. An implementation of the updateFromUIDL() method must include some routine tasks: • Call updateComponent() and return if it succeeds • Manage the component identifier • Manage a reference to the ApplicationConnection object. The widget needs to know it to be able to initiate a server request when a browser event occurs. The latter two of these tasks are not needed if the widget does not handle any user input that needs to be sent to server. The following excerpt provides a skeleton for the updateFromUIDL() method and shows how the component identifier and connection object reference are managed by a widget. String uidlId; ApplicationConnection client; ... public void updateFromUIDL(UIDL uidl, ApplicationConnection client) { if (client.updateComponent(this, uidl, true)) return; this.client = client; uidlId = uidl.getId(); ... }

The updateComponent() call has several functions important for different kinds of components. It updates various default attributes, such as disabled, readonly, invisible, and (CSS) style attributes. If the manageCaption argument is true, the call will also update the caption of the component. By default, the caption is managed by the parent layout of the component. Components, such as a Button, that manage the caption themselves, do not need management of the caption.

230

Deserialization of Component State from Server


Developing Custom Components

The updateComponent() is also part of the transmutation mechanism that allows a single server-side component to have alternative client-side implementations, based on its parameters. For example, the Button server-side component can manifest either as a clickable VButton or as a switchable VCheckBox widget on the client-side. If the parameters are changed, the clientside widget can be replaced with another dynamically. Determination of the correct implementation is done in a WidgetSet. If updateComponent() returns true, the client-side engine can attempt to replace the implementation. For more details on the transmutation mechanism, see Section 10.5, “Defining a Widget Set”. The component identifier is used when the component needs to serialize its updated state to server.The reference to the application connection manager is needed to make the server request. If a component does not have any state changes that need to be sent to the server, management of the variables is not needed. See Section 10.4.2, “Serialization of Component State to Server” below for further details. The design of the client-side framework of Vaadin, because the Paintable is an interface and can not store any references. Having an API layer between GWT and custom widgets would be a much more complicated solution.

10.4.2. Serialization of Component State to Server User input is handled in GWT widgets with events. User input is passed to the server using the updateVariable() method. If the immediate parameter is false, the value is simply added to a queue to be sent to the server at next AJAX request. If the argument is true, the AJAX request is made immediately, and will include all queued updates to variables. The immediate argument is described in more detail below. if (uidl_id == null || client == null) return; client.updateVariable(uidl_id, "myvariable", newvalue, immediate);

The client of the above example is a reference to the ApplicationConnection object that manages server requests. The uidl_id argument is the UIDL identifier obtained during a updateFromUIDL() call with uidl.getId() method. The updateVariable() method has several varieties to send variables of different types.

Table 10.1. UIDL Variable Types Type

Description

UIDL Type

String

String object.

s

int

Native integer value.

i

long

Native long integer value.

l

float

Native single-precision floating-point value.

f

double

Native double-precision floating-point value.

d

boolean

Native boolean value.

b

Object[] Array of object data. The toString() method is used to serialize each a of the objects.The content strings are escaped with escapeString(), to allow characters such as quotes.

Serialization of Component State to Server

231


Developing Custom Components

This serialization mechanism is intended to be as simple as possible in most cases, when the user input is typically just one state variable, while also allowing the serialization of more complex data, if necessary.

Immediateness Server-side components that inherit AbstractComponent have an immediate attribute, set with setImmediate(). This attribute dictates whether a component makes a server request immediately when its state changes, or only afterwards. For example, there is no need to send the contents of a "Username" TextField before the "Login" button has been clicked. On the other hand, the server can set the TextField as immediate to receive changes for example when the component loses focus. Most widgets should support immediateness by receiving the immediate attribute from the UIDL message that renders the widget. The following example is extracted from the VTextField implementation. // Store the immediate attribute in a member variable private boolean immediate = false; ... public void updateFromUIDL(UIDL uidl, ApplicationConnection client) { if(client.updateComponent(this, uidl, true)) return; // Receive and store the immediate attribute immediate = uidl.getBooleanAttribute("immediate"); ... } public void onChange(Widget sender) { if(client != null && id != null) { // Use the stored immediate attribute to say // whether or not make the server request // immediately. client.updateVariable(id, "text", getText(), immediate); } }

In some widgets, the immediate attribute would have little meaning, and in fact an accidental false value would cause undesired behaviour. For example, a button is always expected to send a request to the server when it is clicked. Such widgets can simply use true for the immediate argument in updateVariable(). For example, VButton does as follows: public void onClick(Widget sender) { if (id == null || client == null) return; client.updateVariable(id, "state", true, /* always immediate */ true); }

10.4.3. Example: Integrating the Color Picker Widget Below is a complete example of an integration component for the Color Picker example. It demonstrates all the basic tasks needed for the integration of a GWT widget with its server-side counterpart component. import com.vaadin.terminal.gwt.client.ApplicationConnection; import com.vaadin.terminal.gwt.client.Paintable;

232

Example: Integrating the Color Picker Widget


Developing Custom Components

import com.vaadin.terminal.gwt.client.UIDL; public class VColorPicker extends GwtColorPicker implements Paintable { /** Set the CSS class name to allow styling. */ public static final String CLASSNAME = "example-colorpicker"; /** Component identifier in UIDL communications. */ String uidlId; /** Reference to the server connection object. */ ApplicationConnection client; /** * The constructor should first call super() to initialize * the component and then handle any initialization relevant * to Vaadin. */ public VColorPicker() { // The superclass has a lot of relevant initialization super(); // This method call of the Paintable interface sets // the component style name in DOM tree setStyleName(CLASSNAME); } /** * This method must be implemented to update the client-side * component from UIDL data received from server. * * This method is called when the page is loaded for the * first time, and every time UI changes in the component * are received from the server. */ public void updateFromUIDL(UIDL uidl, ApplicationConnection client) { // This call should be made first. Ensure correct // implementation, and let the containing layout // manage the caption, etc. if (client.updateComponent(this, uidl, true)) return; // Save reference to server connection object to be // able to send user interaction later this.client = client; // Save the UIDL identifier for the component uidlId = uidl.getId(); // Get value received from server and actualize it // in the GWT component setColor(uidl.getStringVariable("colorname")); } /** * Override the method to communicate the new value * to server. **/ public void setColor(String newcolor) { // Ignore if no change if (newcolor.equals(currentcolor.getText())) return; // Let the original implementation to do // whatever it needs to do

Example: Integrating the Color Picker Widget

233


Developing Custom Components

super.setColor(newcolor); // // // if

Updating the state to the server can not be done before the server connection is known, i.e., before updateFromUIDL() has been called. (uidlId == null || client == null) return;

// Communicate the user interaction parameters to server. // This call will initiate an AJAX request to the server. client.updateVariable(uidlId, "colorname", newcolor, true); } }

10.5. Defining a Widget Set The client-side components, or in GWT terminology, widgets, must be made usable in the clientside GWT application by defining a widget set. A widget set is actually a GWT application and needs to be defined in the GWT module descriptor as the entry point of the client-side engine. A GWT module descriptor is an XML file with extension .gwt.xml. If you are using the Eclipse IDE, the New Vaadin Widget wizard will automatically create the GWT module descriptor. See Section 10.2.1, “Creating a Widget” for detailed instructions. The following example of ColorPickerWidgetSet.gwt.xml shows the GWT module descriptor of the Color Picker application. We also define the default stylesheet for the color picker widget, as described above in Section 10.3.3, “Styling GWT Widgets”. <module> <!-- Inherit the default widget set --> <inherits name="com.vaadin.terminal.gwt.DefaultWidgetSet" /> <!-- The default theme of this widget set --> <stylesheet src="colorpicker/styles.css"/> </module>

Compiling widget sets takes considerable time. You can reduce the compilation time significantly by compiling the widget sets only for your browser, which is useful during development. You can do this by setting the user.agent property in the .gwt.xml GWT module descriptor. <set-property name="user.agent" value="gecko1_8"/>

The value attribute should match your browser. The browsers supported by GWT depend on the GWT version, below is a list of browser identifiers supported by GWT 2.0.

Table 10.2. GWT User Agents Identifier Name gecko1_8 Mozilla Firefox 1.5 and later

234

gecko

Mozilla Firefox 1.0 (obsolete)

ie6

Internet Explorer 6

ie8

Internet Explorer 8

safari

Apple Safari and other Webkit-based browsers including Google Chrome

opera

Opera

Defining a Widget Set


Developing Custom Components

For more information about the GWT Module XML Format, please see Google Web Toolkit Developer Guide.

10.6. Server-Side Components Server-side components provide the API for user applications to build their user interface. Many applications do not ever need to bother with the client-side implementation of the standard components, but those that use their own GWT widgets need to have corresponding server-side components. If you are using the Vaadin Plugin for Eclipse, the wizard for creating new widgets will also create a stub of the server-side component for you. See Section 10.2.1, “Creating a Widget” for detailed instructions. A server-side component has two basic tasks: it has to be able to serialize its state variables to the corresponding client-side component, and deserialize any user input received from the client. Many of these tasks are taken care of by the component framework.

10.6.1. Binding to the Client-Side Widget A server-side component needs to be bound to a specific client-side widget. This is done with a special annotation. For example, for the ColorPicker server-side component, we define: import com.vaadin.demo.colorpicker.widgetset.client.ui.VColorPicker; @ClientWidget(VColorPicker.class) public class ColorPicker extends AbstractField { ...

The annotation is read during the compilation of the widget with the GWT Compiler. The compiler invokes a WidgetMapGenerator that reads the annotations from the compiled server-side component classes. It is therefore necessary that the server-side components are compiled before the client-side compilation, as noted in Section 10.8.4, “Compiling GWT Widget Sets”. The serialization is broken down into server-client serialization and client-server deserialization in the following sections. We will also present the complete example of the server-side implementation of the Color Picker component below.

10.6.2. Server-Client Serialization The server-side implementation of a component must be able to serialize its data into a UIDL message that is sent to the client. You need to override the paintContent() method, defined in AbstractComponent. You should call the superclass to allow it to paint its data as well. The data is serialized with the variants of the addAttribute() and addVariable() methods for different basic data types. The UIDL API offered in PaintTarget is covered in Section A.1, “API for Painting Components”.

10.6.3. Client-Server Deserialization The server-side component must be able to receive state changes from the client-side widget. This is done by overriding the changeVariables() method, defined in AbstractComponent. A component should always call the superclass implementation in the beginning to allow it handle its variables.

Server-Side Components

235


Developing Custom Components

The variables are given as objects in the variables map, with the same key with which they were serialized on the client-side. The object type is likewise the same as given for the particular variable in updateVariable() in the client-side. @Override public void changeVariables(Object source, Map variables) { // Let superclass read any common variables. super.changeVariables(source, variables); // Sets the currently selected color if (variables.containsKey("colorname") && !isReadOnly()) { final String newValue = (String)variables.get("colorname"); // Changing the property of the component will // trigger a ValueChangeEvent setValue(newValue, true); } }

The above example handles variable changes for a field component inheriting AbstractField. Fields have their value as the value property of the object. Setting the value with setValue(), as above, will trigger a ValueChangeEvent, which the user of the component can catch with a ValueChangeListener. Contained components, such as components inside a layout, are deserialized by referencing them by their paintable identifier or PID.

10.6.4. Example: Color Picker Server-Side Component The following example provides the complete server-side ColorPicker component for the Color Picker example. It has only one state variable: the currently selected color, which is stored as the property of the component. Implementation of the Property interface is provided in the AbstractField superclass of the component.The UIDL tag name for the component is colorpicker and the state is communicated through the colorname variable. package com.vaadin.demo.colorpicker; import com.vaadin.demo.colorpicker.widgetset.client.ui.VColorPicker; ... /** * Color picker for selecting a color from a palette. * * @author magi */ @ClientWidget(VColorPicker.class) public class ColorPicker extends AbstractField { public ColorPicker() { super(); setValue(new String("white")); } /** The property value of the field is a String. */ @Override public Class<?> getType() { return String.class; } /** Set the currently selected color. */ public void setColor(String newcolor) { // Sets the color name as the property of the component. // Setting the property will automatically cause // repainting of the component with paintContent().

236

Example: Color Picker Server-Side Component


Developing Custom Components

setValue(newcolor); } /** Retrieve the currently selected color. */ public String getColor() { return (String) getValue(); } /** Paint (serialize) the component for the client. */ @Override public void paintContent(PaintTarget target) throws PaintException { // Superclass writes any common attributes in the // paint target. super.paintContent(target); // Add the currently selected color as a variable in // the paint target. target.addVariable(this, "colorname", getColor()); } /** Deserialize changes received from the client. */ @Override public void changeVariables(Object source, Map variables) { // Sets the currently selected color if (variables.containsKey("colorname") && !isReadOnly()) { // Changing the property of the component will // trigger a ValueChangeEvent setValue((String) variables.get("colorname"), true); } } }

10.7. Using a Custom Component A custom component is used like any other Vaadin component. You will, however, need to compile the client-side widget set with the GWT Compiler. See Section 10.8.4, â&#x20AC;&#x153;Compiling GWT Widget Setsâ&#x20AC;? for instructions on how to compile widget sets.

10.7.1. Example: Color Picker Application The following server-side example application shows how to use the Color Picker custom widget. The example includes also server-side feedback of the user input and changing the color selection to show that the communication of the component state works in both directions. package com.vaadin.demo.colorpicker; import import import import

com.vaadin.data.Property.ValueChangeEvent; com.vaadin.data.Property.ValueChangeListener; com.vaadin.ui.*; com.vaadin.ui.Button.ClickEvent;

/** * Demonstration application that shows how to use a simple * custom client-side GWT component, the ColorPicker. */ public class ColorPickerApplication extends com.vaadin.Application { Window main = new Window("Color Picker Demo"); /* The custom component. */ ColorPicker colorselector = new ColorPicker();

Using a Custom Component

237


Developing Custom Components

/* Another component. */ Label colorname; public void init() { setMainWindow(main); setTheme("demo"); // Listen for value change events in the custom // component, triggered when user clicks a button // to select another color. colorselector.addListener(new ValueChangeListener() { public void valueChange(ValueChangeEvent event) { // Provide some server-side feedback colorname.setValue("Selected color: " + colorselector.getColor()); } }); main.addComponent(colorselector); // Add another component to give feedback from // server-side code colorname = new Label("Selected color: " + colorselector.getColor()); main.addComponent(colorname); // Server-side manipulation of the component state Button button = new Button("Set to white"); button.addListener(new Button.ClickListener() { public void buttonClick(ClickEvent event) { colorselector.setColor("white"); } }); main.addComponent(button); } }

10.7.2. Web Application Deployment Deployment of web applications that include custom components is almost identical to the normal case where you use only the default widget set of Vaadin. The default case is documented in Section 4.8.3, â&#x20AC;&#x153;Deployment Descriptor web.xmlâ&#x20AC;?. You only need to specify the widget set for the application in the WebContent/WEB-INF/web.xml deployment descriptor. If you use the Vaadin Plugin for Eclipse to create a new widget in your project, the plugin will modify the deployment descriptor to use the custom widget set. The following deployment descriptor specifies the Color Picker Application detailed in the previous section. <?xml version="1.0" encoding="UTF-8"?> <web-app id="WebApp_ID" version="2.4" xmlns="http://java.sun.com/xml/ns/j2ee" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://java.sun.com/xml/ns/j2ee http://java.sun.com/xml/ns/j2ee/web-app_2_4.xsd"> <display-name>myproject</display-name> <servlet> <servlet-name>ColorPickerServlet</servlet-name> <servlet-class> com.vaadin.terminal.gwt.server.ApplicationServlet

238

Web Application Deployment


Developing Custom Components

</servlet-class> <init-param> <param-name>application</param-name> <param-value> com.vaadin.demo.colorpicker.ColorPickerApplication </param-value> </init-param> <init-param> <param-name>widgetset</param-name> <param-value> com.vaadin.demo.colorpicker.widgetset.ColorPickerWidgetSet </param-value> </init-param> </servlet> <servlet-mapping> <servlet-name>ColorPickerServlet</servlet-name> <url-pattern>/*</url-pattern> </servlet-mapping> </web-app>

The project specific parameters are emphasized. Notice that the widget set name is not a file name, but the base name for the ColorPickerWidgetSet.gwt.xml module descriptor. The Eclipse Plugin will automatically generate the init-param parameter in the web.xml file of your project when you create a new widget. As the project context root in the above example is myproject and the <url-pattern> is /*, the URL for the application will be /myproject/. If you are using an URL pattern such as /myapp/*, you need to make an additional mapping to map requests to /VAADIN/* context to the same servlet. Otherwise the default widget set and built-in themes in Vaadin will be missing. <servlet-mapping> <servlet-name>Book of Vaadin Examples</servlet-name> <url-pattern>/VAADIN/*</url-pattern> </servlet-mapping>

10.8. GWT Widget Development Development of new GWT widgets includes widget set definition (GWT Module Descriptor), compiling the widgets and the Vaadin Client-Side Engine to JavaScript with the GWT Compiler, and debugging the application in the GWT Development Mode. You can use any IDE for developing GWT components for Vaadin. The examples given in this book are for the Eclipse IDE. It allows easy launching of the GWT Development Mode, debugging, and running an external compiler for GWT widget sets.

10.8.1. Creating a Widget Project Creation of a Vaadin project that uses the default widget set was covered in Section 2.4, â&#x20AC;&#x153;Your First Project with Vaadinâ&#x20AC;?. Developing custom widgets creates a number of additional requirements for a project. Let us review the steps required for creating an application that contains custom widgets. Details for each step are given in the subsequent sections. 1. Create a new Vaadin project (for Eclipse, see Section 2.4.1)

GWT Widget Development

239


Developing Custom Components

2. Copy or import the GWT JARs to the project and add them to the class path (Section 10.8.2) 3. Write the source code for the client-side widgets, their server-side counterparts, and the application (Section 10.8.3) 4. Write the web.xml Deployment Descriptor for the web application • Define the widget set used in the application (Section 10.7.2) 5. Compile the widget set to JavaScript runtime with GWT Compiler (Section 10.8.4) 6. Deploy the project to an application server (for Eclipse and Tomcat see Section 2.4.3) 7. Either: a. Open a web browser to use the web application. b. Open a web browser in GWT Development Mode to debug the client-side widget code. (Section 10.8.6) The contents of a ready widget development project are described in Section 10.8.5, “Ready to Run”. The Vaadin Plugin for Eclipse makes the creation of application custom widgets as well as the application projects easy by taking care of all the above steps (except deployment), so if you use it, please see Section 10.2, “Doing It the Simple Way in Eclipse” for detailed instructions.

10.8.2. Importing GWT Installation Package You will need to include the Google Web Toolkit in your project to develop custom widgets. The installation directory of Vaadin includes full GWT installation in the gwt subdirectory. The package includes precompiled libraries and applications for the specific platform of the installation. To use the libraries, you need to configure them in the classpath of your project as described below. You can copy or import either the entire gwt directory or just the GWT JARs to your project. If you use Eclipse and copy the GWT directory or the libraries to the project with system tools, remember to select your project folder and press F5 to refresh the project. You can also import the directory as follows (importing the JARs is similar): 1. Right-click on the project folder in Project Explorer and select Import → Import.... 2. From the Import dialog, select General → File System and click Next. 3. Click Browse button of the "From directory" field and browse to the gwt directory under the Vaadin installation directory. Click Ok in the file selection dialog. 4. Select the gwt entry in the list box for importing. 5. In the "Into folder" field, enter myproject/gwt. (If you do not set this, all the contents of the gwt directory will be imported directly below the root directory of the project which is undesirable.) 6. Click Finish.

240

Importing GWT Installation Package


Developing Custom Components

You must include the GWT libraries in the classpath of the project. If using Eclipse, right-click on the project folder in the Project Explorer in Eclipse and select Properties. Select Java Build Path → Libraries.

10.8.3. Writing the Code Guidelines for the Project Structure The project structure is otherwise free, but if you use the build script described in Section 10.8.4, “Compiling GWT Widget Sets” or configure the GWT Development Mode, as described in Section 10.8.6, “GWT Development Mode”, the examples assume that source files are located under the WebContent/WEB-INF/src folder. We recommend that the name for the package containing the widget set is widgetset, but this is not required. GWT does, however, require that the client-side code is written under a client package and any stylesheets are located under a public directory (not necessarily a package) under the package containing the .gwt.xml descriptor. See Section 10.8.5, “Ready to Run” for an example project structure.

Importing the ColorPicker Demo If you want to use the Color Picker application as an application skeleton, you need to import it under the source folder. 1. Right-click on the source folder and select Import. 2. In the Import dialog, select General → File System and click Next. 3. Browse to WebContent/WEB-INF/src/com/vaadin/demo/colorpicker/ and click Ok button in the Import from directory dialog. 4. I n t h e I n t o f o l d e r f i e l d , e n t e r myproject/WebContent/WEB-INF/src/com/vaadin/demo/colorpicker. 5. Check the colorpicker entry in the list box. 6. Click Finish. This will import the directory as com.vaadin.demo.colorpicker package. If you want to use it as a skeleton for your own project, you should refactor it to some other name. Notice that you will need to refactor the package and application name manually in the web.xml and .gwt.xml descriptor files.

10.8.4. Compiling GWT Widget Sets You need to compile the Vaadin Client-Side Engine and your custom widget set to JavaScript with the GWT Compiler. Vaadin installation package includes an Ant build script build-widgetset.xml in the WebContent/docs/example-source/ directory. If you are using the Vaadin Plugin for Eclipse, it will enable a toolbar button that allows you to compile the widget set. See Section 10.2.2, “Recompiling the Widget Set” for instructions. To compile the Color Picker widget set example using the Ant build script, just change to the directory and enter:

Writing the Code

241


Developing Custom Components

$ ant -f build-widgetset.xml

We advice that you copy the build script to your project and use it as a template. Just set the paths in the "configure" target and the widget set class name in the "compile-widgetset" target to suit your project. Alternatively, you can launch the build script from Eclipse, by right-clicking the script in Package Explorer and selecting Run As → Ant Build. Progress of the compilation is shown in the Console window. After compilation, refresh the project by selecting it and pressing F5. This makes Eclipse scan new content and become aware of the output of the compilation in the WebContent/VAADIN/widgetsets/ directory. If the project is not refreshed, the JavaScript runtime is not included in the web application and running the application will result in an error message such as the following: Requested resource [VAADIN/widgetsets/com.vaadin.demo.colorpicker.gwt.ColorPickerWidgetSet/com.vaadin.demo.colorpicker.gwt.ColorPickerWidgetSet.nocache.js] not found from filesystem or through class loader. Add widgetset and/or theme JAR to your classpath or add files to WebContent/VAADIN folder.

Compilation with GWT is required also initially when using the GWT Development Mode, which is described in Section 10.8.6, “GWT Development Mode”.The compilation with the GWT Compiler must be done at least once, as it provides files that are used also by the GWT Development Mode, even though the browser runs the GWT application in Java Virtual Machine instead of JavaScript. Warning Because GWT supports a slightly reduced version of Java, GWT compilation can produce errors that do not occur with the Java compiler integrated in the Eclipse IDE. Also notice that client-side compilation loads the server-side classes (to find the @ClientWidget annotations) and, as a side effect, executes any static code blocks in the classes, even in any non-component classes such as the application class. This could cause unexpected behaviour during the compilation.

Compiling a Custom Widget Set If you wish to use the build script to compile your own widget sets, open it in an editor. The build script contains some instructions in the beginning of the file. First, you need to make some basic configuration in the configure target: <target name="configure"> <!-- Path from this file to the project root --> <property name="base" value="../../../" /> <!-- Location of GWT distribution --> <property name="gwt-location" value="${base}gwt" /> <!-- Location of Vaadin JAR --> <property name="vaadin-jar-location" value="${base}WebContent/WEB-INF/lib/vaadin-6.3.3.jar"/> <!-- Location of project source code --> <property name="src-location"

242

Compiling GWT Widget Sets


Developing Custom Components

value="${base}WebContent/WEB-INF/src" /> <!-- Target where to compile server-side classes --> <property name="server-side-destination" value="${base}WebContent/WEB-INF/classes"/> <!-- Target where to compile the widget set --> <property name="client-side-destination" value="${base}WebContent/VAADIN/widgetsets" /> </target>

You also need to define the widget set in the configure-widgetset target: <!-- NOTE: Modify this example to compile your own widgetset --> <target name="configure-widgetset"> <!-- Name of the widget set --> <property name="widgetset" value="com.vaadin.demo.colorpicker.gwt.ColorPickerWidgetSet"/>

The widget set class name must match the package and the file name of the .gwt.xml descriptor, without the extension. You can use the compile-widgetset target as it is, or as a template for your own configuration: <!-- Build the widget set. --> <target name="compile-widgetset" depends="compile-server-side, generate-widgetset"> <echo>Compiling ${widgetset}...</echo> <java classname="com.google.gwt.dev.Compiler" failonerror="yes" fork="yes" maxmemory="256m"> <arg value="-war" /> <arg value="${client-side-destination}" /> <arg value="${widgetset}" /> <jvmarg value="-Xss1024k"/> <jvmarg value="-Djava.awt.headless=true"/> <classpath> <path refid="compile.classpath"/> </classpath> </java> </target>

Notice that the server-side must be compiled before the client-side and the compiled server-side classes must be included in the class path for the GWT Compiler. The reason is that the annotations that define the component-to-widget mappings are read from the class files during the GWT compilation. Google Web Toolkit Version You should use a version of GWT suitable for the version of Vaadin you are using; we recommend using the GWT included in the installation package, but other versions may work as well. If you are upgrading from Vaadin 5, the GWT 1.6 and later versions used by Vaadin 6 contains a new compiler and the old GWTCompiler class used previously for compiling GWT widgets is deprecated and replaced with com.google.gwt.dev.Compiler. You should update your existing widget set build scripts or launch configurations to use the new compiler class. The only significant API change is the output directory parameter, previously -out, now -war, as shown in the example above.

Compiling GWT Widget Sets

243


Developing Custom Components

Java Stack Overflow Problems The -Xss parameter for the Java process may be necessary if you experience stack overflow errors with the default stack size. They occur especially with GWT 1.6, which uses large amount of stack space. Notice further that the Windows version of Sun JRE 1.5 has a bug that makes the stack size setting ineffective. The Windows version also has a smaller default stack size than the other platforms. If you experience the problem, we advice that you either use JRE 1.6 on the Windows platform or use a wrapper that circumvents the problem. To use the wrapper, use class com.vaadin.launcher.WidgetsetCompiler in the build script instead of the regular compiler. The -Djava.awt.headless=true is relevant in Linux/UNIX platforms to avoid some X11 warnings. You can now compile the widget set with the following command: $ ant -f build-widgetset.xml

Generating Widget Set Definition If you use libraries containing other widget sets, you may want to automatically generate the .gwt.xml GWT Module Descriptor that defines the widget set that combines the Vaadin default widget set, widget sets included in any libraries, and any widget sets defined in your project. If you use the build-widgetset.xml Ant build script as a template, you need to define the settings required for generating a widget set in the configure-widgetset target. <target name="configure-widgetset"> <property name="widgetset" value="com.vaadin.demo.generated.GeneratedWidgetSet"/> <property name="generate.widgetset" value="1"/> <property name="widgetset-path" value="com/vaadin/demo/widgetset"/> </target>

Define the name of the widget set definition file, without the .gwt.xml suffix, in the widgetset property. The generate-widgetset target is executed only if the generate.widgetset property is defined. You also need to give a file path to the widget set directory, relative to the $src-location directory.. The generate-widgetset target is defined as follows: <target name="generate-widgetset" depends="compile-server-side, configure-widgetset" if="generate.widgetset"> <!-- Create the directory if it does not already exist. --> <mkdir dir="${src-location}/${widgetset-path}"/> <java classname="com.vaadin.terminal.gwt.widgetsetutils.WidgetSetBuilder" failonerror="yes" fork="yes" maxmemory="256m"> <arg value="${widgetset}"/> <jvmarg value="-Xss1024k"/> <jvmarg value="-Djava.awt.headless=true"/> <classpath> <path refid="compile.classpath"/> </classpath> </java> </target>

244

Compiling GWT Widget Sets


Developing Custom Components

The widget set builder assumes that the root of the source directory is the first item in the class path, here defined with the src-location property. The location must point to the root of the entire source tree, not the path of the widget set under it. Also the location of compiled serverside classes must be included, here defined with the server-side-destination property. In the example script, the class path is defined in the init target as follows: <path id="compile.classpath"> <pathelement path="${src-location}" /> <pathelement path="${server-side-destination}" /> <pathelement path="${toolkit-jar-location}" /> <pathelement path="${gwt-location}/gwt-user.jar" /> <pathelement path="${gwt-location}/gwt-dev-${gwt-platform}.jar" /> <fileset dir="${base}WebContent/WEB-INF/lib/"> <include name="**/*.jar"/> </fileset> </path>

10.8.5. Ready to Run Figure 10.8, â&#x20AC;&#x153;Annotated Project Contentsâ&#x20AC;? shows the contents of a ready project.

Figure 10.8. Annotated Project Contents

Ready to Run

245


Developing Custom Components

Notice that the Package Explorer does not correspond with the file system contents. Eclipse displays the items marked with asterisk (*) in a logical location, instead of the physical location in the file system. You can either run the application in web mode, as introduced in Section 2.4.4, or debug it in the GWT Development Mode, as detailed in the next section.

10.8.6. GWT Development Mode The GWT Development Mode allows debugging client-side GWT applications in a Java IDE such as Eclipse, as if the applications ran as Java in the browser, even though they actually are JavaScript. This is made possible by the Google Web Toolkit Developer Plugin, which is available for major browsers, such as Mozilla Firefox, Google Chrome, Safari, and Internet Explorer.

Figure 10.9. GWT Development Mode

Figure 10.9, “GWT Development Mode” shows the GWT Development Mode in action. On the left, you have the GWT Development Mode window. It displays compilation information and possible errors that occur during compilation. You can open the application in a new browser window by clicking Launch Default Browser. The browser window has a Compile/Browse button, which runs the GWT Compiler to produce the runtime code and opens a regular web browser to run the application. Notice that even though it is possible to recompile the program with the button, GWT Compiler must be run at least once before launching the GWT Development Mode, as described in Section 10.8.4, “Compiling GWT Widget Sets”, as the compiler creates some required files. Because GWT supports a slightly reduced version of Java, GWT compilation can produce errors that do not occur with the Java compiler integrated in the Eclipse IDE. Such errors will show up in the GWT Development Mode window.

246

GWT Development Mode


Developing Custom Components

Configuring GWT Development Mode Launching in Eclipse This section gives details on creating a launch configuration for the GWT Development Mode in the Eclipse IDE. We use the QuickStart installation of Vaadin covered in Section 2.3, “QuickStart with Eclipse” as an example project. The project includes source code for the Color Picker demo application. If you use the Vaadin Plugin for Eclipse, it can create a launch configuration for launching the GWT Development Mode. See Section 10.2.3, “Plugin Related Project Settings” for instructions. 1. Select from menu Run → Debug... and the Debug configuration window will open. Notice that it is not purposeful to run the GWT Development Mode in the "Run" mode, because its entire purpose is to allow debugging. 2. Select the Java Application folder and click on the New button to create a new launch configuration.

Figure 10.10. Creating New Launch Configuration

3. Click on the created launch configuration to open it on the right-side panel. In the Main tab, give the launch configuration a name. Define the Main class as com.google.gwt.dev.DevMode.

GWT Development Mode

247


Developing Custom Components

Figure 10.11. Naming Launch Configuration

4. Switch to the Arguments tab and enter arguments for the application. a. In the Program arguments field, enter: -noserver -war WebContent/VAADIN/widgetsets com.example.myproject.widgetset.MyProjectWidgetSet -startupUrl http://localhost:8080/myproject

The browser application, DevMode, takes as its arguments the following parameters:

248

-noserver

Prevents an embedded web server from starting, thereby allowing to use an already running server.

-whitelist

Adds a regular expression to the list of allowed URL patterns for the web browser. Modify the port number from the 8080 given above as necessary.

-war

Output directory for compiling widgets with GWT Compiler. T h e d i r e c t o r y m u s t b e WebContent/VAADIN/widgetsets. You can compile the widgets either from the GWT Development Mode window or externally as explained later in this chapter.

-startupUrl <URL>

The URL to connect to. This must be the same as the whitelist entry given above. The port number must correspond to the port of the running web server. The Jetty web server included in Vaadin will run in port 8888 by default. In contrast, Apache Tomcat installed under Eclipse will run in port 8080 by default.

GWT Development Mode


Developing Custom Components

b. In the VM arguments field enter, for example, -Xms256M -Xmx512M to give the GWT Development Mode more memory than the default amount. On Mac, add also -XstartOnFirstThread.

Figure 10.12. DevMode Arguments

5. In the Classpath tab, you will by default have vaadin-examples, which contains the default classpath entries for the project. If the classpath entries for the project are sufficient, this should be enough. 6. Click Apply to save the launch configuration. 7. Click Debug to launch the GWT Development Mode using the launch configuration. See the following section for details on debugging with the GWT Development Mode.

Debugging with GWT Development Mode The purpose of the GWT Development Mode is to allow debugging client-side GWT applications, or in our case, GWT widgets. Below is a checklist for important requirements for launching the GWT Development Mode: • GWT is installed in the project. • GWT libraries are included in the project classpath. • Widget sets have been compiled with GWT Compiler. • web.xml descriptor is configured. • Web server is running and listening to the correct port.

GWT Development Mode

249


Developing Custom Components

â&#x20AC;˘ GWT Development Mode launch configuration is configured. Once everything is ready to start debugging, just open a source file, for example, the com.vaadin.demo.colorpicker.gwt.client.ui.GwtColorPicker class. Find the onClick() method. At the line containing the setColor() call, right-click on the leftmost bar in the editor and select Toggle Breakpoint from the popup menu. A small magnifying glass will appear in the bar to indicate the breakpoint.

Figure 10.13. Setting a Breakpoint

Select from menu Run â&#x2020;&#x2019; Debug... and the Debug configuration window will open. Notice that it is not purposeful to run the GWT Development Mode in the "Run" mode, because its entire purpose is to allow debugging.

Figure 10.14. Debugging with GWT Development Mode

Starting demo applications under the GWT Development Mode can take considerable time! Compiling widgets can take 10-60 seconds, depending on the hardware. During this time, the web browser is unresponsive, does not update its window, and appears "stuck".

250

GWT Development Mode


Developing Custom Components

Please refer to Eclipse IDE documentation for further instructions on using the debugger.

10.8.7. Packaging a Widget Set Packaging and reusing custom components is almost as easy as with any Java code, but with a small difference. You can package the source code of a widget set as a JAR. A user can drop the JAR in the project and add it to the class path. If using widgets from multiple widget sets, the user has to inherit the widget set in a combining widget set (which can be generated automatically). The user then has to compile the widget set(s) as described in Section 10.8.4, “Compiling GWT Widget Sets”. While you could, in theory, package precompiled widget sets in the JAR as well, it would create a serious danger for incompatibility between the client-side and server-side versions of Vaadin. The user would anyhow have to recompile the widget set if he uses multiple widget sets packaged in different JARs, or has developed project-specific widgets. The build-widgetset.xml Ant script in the WebContent/docs/example-source directory of the Vaadin installation package includes an example package-jar target for building a JAR. You can use the example as it is or modify it as you need. You need to make the JAR packaging specific configuration in the configure-jar target. Change to property values to reflect your widget set. <target name="configure-jar"> <!-- The compiled JAR name --> <property name="jar-destination" value="${base}colorpicker.jar"/> <!-- Title of the widget set (for JAR) --> <property name="widgetset-title" value="ColorPicker"/> <!-- Version of the widget set (for JAR) --> <property name="widgetset-version" value="1.0"/> <!-- Vendor of the widget set (for JAR) --> <property name="widgetset-vendor" value="IT Mill Oy"/> </target>

You may want to check also the package-jar target if you want to use other license information or otherwise customize the package content. Assuming that you have otherwise configured the build script for your project as described in Section 10.8.4, “Compiling GWT Widget Sets”, you can build the JAR package with the following command: $ ant -f build-widgetset.xml package-jar

Notice that the package-jar target does not depend on the compile-widgetset target, as the compiled widget set is not included in the package. If you really wish to do so, add the dependency and include the compiled files as a fileset for the package.

10.8.8. Troubleshooting Below are some typical cases that may cause you trouble. Deferred binding fails

The OOPHM widget set compilation produces the following error in Eclipse console:

Packaging a Widget Set

251


Developing Custom Components

Deferred binding failed for 'com.vaadin.csvalidation.widgetset.client.ui.VJavaScriptEditor' (did you forget to inherit a required module?)

This problem occurs if the creation of an instance of a widget fails, usually due to an exception in the constructor. The GWT Development Mode console displays a more detailed exception log. Widgets are missing from the widget set

The widget set compilation displays the list of widgets included in the widget set. If the list includes only the inherited widgets (built-in Vaadin widgets), but your custom widgets are missing, there is a problem. The typical reason is that the server-side classes are not yet compiled or are missing from the classpath. Check that the server-side compilation output folder (such as build/classes) is included in the classpath of GWT Compiler and that you compile the server-side components before the client-side, so that they are available in the classpath. The reason for this is that GWT Compiler (or more precisely a Vaadin widget set generator) reads the annotations that define the mappping from server-side components to client-side widgets from compiled class files.

The correct widget set is not loaded

The widget set must be specified in WebContent/WEB-INF/web.xml descriptor, as an initialization parameter for the servlet. <init-param> <description>Application widgetset</description> <param-name>widgetset</param-name> <param-value>com.example.mylibrary.widgetset.MyWidgetset</param-value> </init-param>

A generated widget set inherits itself

252

The widget set builder can create a .gwt.xml descriptor that inherits the same widget set.This usually occurs when the class path for the source files is set to something different than the root of the entire source tree, so that the package name of the widget set to be defined with the descriptor would be invalid.

Troubleshooting


Chapter 11

Advanced Web Application Topics 11.1. Special Characteristics of AJAX Applications ...................................... 254 11.2. Application-Level Windows .................................................................. 254 11.3. Embedding Applications in Web Pages ............................................... 261 11.4. Debug and Production Mode ............................................................... 266 11.5. Resources ........................................................................................... 268 11.6. Shortcut Keys ...................................................................................... 271 11.7. Printing ................................................................................................ 275 11.8. Portal Integration ................................................................................. 276 11.9. Google App Engine Integration ........................................................... 291 11.10. Common Security Issues ................................................................... 292 11.11. URI Fragment and History Management with UriFragmentUtility ................................................................................................................. 293 11.12. Capturing HTTP Requests ................................................................ 294 11.13. Drag and Drop ................................................................................... 297 11.14. Using Add-on Components ............................................................... 305 This chapter covers various features and topics often needed in applications. While other topics could be considered as "advanced", the first section gives a brief introduction to AJAX development for beginners.

Book of Vaadin

253


Advanced Web Application Topics

11.1. Special Characteristics of AJAX Applications New to AJAX? This section is intended for people familiar with the development of either traditional web applications or desktop applications, who are entering AJAX-enabled web application development. AJAX application development has a few special characteristics with respect to other types of applications. Possibly the most important one is how the display is managed in the web browser. The web was originally not built for applications, but for hypertext pages that you can view with a browser. The purpose of web pages is to provide content for the user. Application software has a somewhat different purpose; usually to allow you to work on some data or content, much of which is not ever intended to be accessible through a web browser as web pages. As the web is inherently page-based, conventional web applications had to work with page requests and output HTML as response. JavaScript and AJAX have made it possible to let go of the pages. Pages are largely an unknown concept to conventional desktop applications. At most, desktop applications can open multiple windows, but usually they work with a single main window, with an occasional dialog window here and there. Same goes usually for web applications developed with Vaadin: an application typically runs on a single page, changing the layout as needed and popping up dialog boxes. Not having to load pages and use hyperlinks to communicate all user interaction is a relief for application development. However, they are an important feature that ordinary desktop applications lack. They allow referencing different functionalities of an application or resources managed by the application. They are also important for integration with external applications. Certain resources can be identified through a URI or Universal Resource Identifier. A URI can easily be passed around or stored as a bookmark. We will see in Section 11.5.1, “URI Handlers” how you can retrieve the URI of a page request. Similarly, a page request can have query parameters, which can be handled as detailed in Section 11.5.2, “Parameter Handlers”. Using URIs or request parameters to access functionalities or content is not as straight-forward as in conventional page-based web applications. Vaadin, just as any other AJAX framework, uses browser cookies not just for tracking users but also for tracking the application state. Cookies are unique in a browser, so any two windows share the same cookies and therefore also the state. The advantage is that you can close your browser and open it again and the application will be in the state where you left off (except for components such as text fields which did not have the immediate attribute enabled). The disadvantage is that there is no good way to distinguish between the windows, so there can usually be only a single application window. Even if there were several, you would have trouble with synchronization of application data between windows. Many conventional page-based web applications simply ignore out-of-sync situations, but such situations are risky for application platforms that are intended to be stable. Therefore it is safer to work with a single browser window. If you wish to have multiple windows in your application, you can create them inside the main window as Window objects. A URI can be used to fetch resources that have no particular state or to provide an entry point to the application.

11.2. Application-Level Windows Vaadin Release 5 introduces support for multiple application-level windows that can be used just like the main window. All such windows use the same application session. Each window is identified with a URL that is used to access it. This makes it possible to bookmark applicationlevel windows. Such windows can even be created dynamically based on URLs.

254

Special Characteristics of AJAX Applications


Advanced Web Application Topics

Application-level windows allow several uses important for the usability of browser-based applications. • Native child windows. An application can open child windows that are not floating windows inside a parent window. • Page-based browsing. The application can allow the user to open certain content to different windows. For example, in a messaging application, it can be useful to open different messages to different windows so that the user can browse through them while writing a new message. • Bookmarking. Bookmarks in the web browser can provide an entry-point to some content provided by an application. • Embedding windows. Windows can be embedded in web pages, thus making it possible to provide different views to an application from different pages or even from the same page, while keeping the same session. See Section 11.3, “Embedding Applications in Web Pages”. Because of the special nature of AJAX applications, these uses require some caveats. We will go through them later in Section 11.2.4, “Caveats in Using Multiple Windows”.

11.2.1. Creating New Application-Level Windows Creating a new application-level window is much like creating a child window (see Section 4.3, “Child Windows”), except that the window is added with addWindow() to the application object instead of the main window. public class WindowTestApplication extends Application { public void init() { // First create the main window. final Window main = new Window ("My Test Application"); setMainWindow(main); // Create another application-level window. final Window mywindow = new Window("Second Window"); // Manually set the name of the window. mywindow.setName("mywindow"); // Add some content to the window. mywindow.addComponent(new Label("Has content.")); // Add the window to the application. addWindow(mywindow); } }

This creates the window object that a user can view by opening a URL in a browser. Creating an application-level window object does not open a new browser window automatically to view the object, but if you wish to open one, you have to do it explicitly as shown below. An applicationlevel window has a unique URL, which is based on the application URL and the name of the window given with the setName() method. For example, if the application URL is http://localhost:8080/myapp/ and the window name is mywindow, the URL for the window will be http://localhost:8080/myapp/mywindow/. If the name of a window is not explicitly set with setName(), an automatically generated name will be used. The name can be retrieved with the getName() method and the entire URL with getURL().

Creating New Application-Level Windows

255


Advanced Web Application Topics

There are three typical ways to open a new window: using the open() method of Window class, a Link, or referencing it from HTML or JavaScript code written inside a Label component. The Window open() method takes as parameters a resource to open and the target name.You can use ExternalResource to open a specific URL, which you get from the window to be opened with the getURL() method. /* Create a button to open a new window. */ main.addComponent(new Button("Click to open new window", new Button.ClickListener() { public void buttonClick(ClickEvent event) { // Open the window. main.open(new ExternalResource(mywindow.getURL()), "_new"); } }));

The target name is one of the default HTML target names (_new, _blank, _top, etc.) or a custom target name. How the window is exactly opened depends on the browser. Browsers that support tabbed browsing can open the window in another tab, depending on the browser settings. Another typical way to open windows is to use a Link component with the window URL as an ExternalResource. /* Add a link to the second window. */ Link link = new Link("Click to open second window", new ExternalResource(mywindow.getURL())); link.setTargetName("second"); link.setTargetHeight(300); link.setTargetWidth(300); link.setTargetBorder(Link.TARGET_BORDER_DEFAULT); main.addComponent(link);

Using a Link allows you to specify parameters for the window that opens by clicking on the link. Above, we set the dimensions of the window and specify what window controls the window should contain. The Link.TARGET_BORDER_DEFAULT specifies to use the default, which includes most of the usual window controls, such as the menu, the toolbar, and the status bar. Another way to allow the user to open a window is to insert the URL in HTML code inside a Label. This allows even more flexibility in specifying how the window should be opened. // Add the link manually inside a Label. main.addComponent( new Label("Second window: <a href='" + mywindow.getURL() + "' target='second'>click to open</a>", Label.CONTENT_XHTML)); main.addComponent( new Label("The second window can be accessed through URL: " + mywindow.getURL()));

When an application-level window is closed in the browser the close() method is normally called just like for a child window and the Window object is purged from the application. However, there are situations where close() might not be called. See Section 11.2.3, â&#x20AC;&#x153;Closing Windowsâ&#x20AC;? for more information.

11.2.2. Creating Windows Dynamically You can create a window object dynamically by its URL path by overriding the getWindow() method of the Application class. The method gets a window name as its parameter and must return the corresponding Window object. The window name is determined from the first URL

256

Creating Windows Dynamically


Advanced Web Application Topics

path element after the application URL (window name may not contain slashes). See the notes below for setting the actual name of the dynamically created windows below. The following example allows opening windows with a window name that begins with "planet-" prefix. Since the method is called for every browser request for the application, we filter only the requests where a window with the given name does not yet exist. public class WindowTestApplication extends Application { ... @Override public Window getWindow(String name) { // If a dynamically created window is requested, but // it does not exist yet, create it. if (name.startsWith("planet-") && super.getWindow(name) == null) { String planetName = name.substring("planet-".length()); // Create the window object. Window newWindow = new Window("Window about " + planetName); // DANGEROUS: Set the name explicitly. Otherwise, // an automatically generated name is used, which // is usually safer. newWindow.setName(name); // Put some content in it. newWindow.addComponent( new Label("This window contains details about " + planetName + ".")); // Add it to the application as a regular // application-level window. addWindow(newWindow); return newWindow; } // Otherwise the Application object manages existing // windows by their name. return super.getWindow(name); }

The window name is and must be a unique indentifier for each Window object instance. If you use setName() to set the window name explicitly, as we did above, any browser window that has the same URL (within the same browser) would open the same window object. This is dangerous and generally not recommended, because the browser windows would share the same window object. Opening two windows with the same static name would immediately lead to a synchronization error, as is shown in Figure 11.1, â&#x20AC;&#x153;Synchronization Error Between Windows with the Same Nameâ&#x20AC;? below. (While also the window captions are same, they are irrelevant for this problem.)

Creating Windows Dynamically

257


Advanced Web Application Topics

Figure 11.1. Synchronization Error Between Windows with the Same Name

There are some cases where setting the name explicitly is useful. The launch application below is one example, as it always opens the other windows in a window target that is specific to the window name, thereby never creating two windows with the same URL. Similarly, if you had embedded the application in a browser frame and the link would open the window in a frame, you would not have problems. Having a single window instance for a URL is also useful if the browser crashes and the user opens the window again, as it will have kept its previous (serverside) state. Leaving the window name to be automatically generated allows opening multiple windows with the same URL, while each of the windows will have a separate state. The URL in the location bar stays unchanged and the generated window name is used only for the Ajax communications to identify the window object. A generated name is a string representation of a unique random number, such as "1928676448". You should be aware of the generated window names when overriding the getWindow() method (and not unintentionally create a new window instance dynamically for each such request). The condition in the above example would also filter out the requests for an already existing window with a generated name. Figure 11.2, “A Dynamically Created Window” shows a dynamically created application-level window with the URL shown in the address bar. The URL for the application is here http://localhost:8080/tk5/windowexample/, including the application context, and the dynamically created window's name is planet-mars.

Figure 11.2. A Dynamically Created Window

The application knows the windows it already has and can return them after the creation. The application also handles closing and destruction of application-level window objects, as discussed in Section 11.2.3, “Closing Windows”. Such dynamic windows could be opened as in the following example:

258

Creating Windows Dynamically


Advanced Web Application Topics

public void init() { final Window main = new Window("Window Test"); setMainWindow(main); // Have some IDs for the dynamically creatable windows. final String[] items = new String[] { "mercury", "venus", "earth", "mars", "jupiter", "saturn", "uranus", "neptune" }; // Create a list of links to each of the available window. for (int i = 0; i < items.length; i++) { // Create a URL for the window. String windowUrl = getURL() + "planet-" + items[i]; // Create a link to the window URL. Using the // item ID for the target also opens it in a new // browser window (or tab) unique to the window name. main.addComponent( new Link("Open window about " + items[i], new ExternalResource(windowUrl), items[i], -1, -1, Window.BORDER_DEFAULT)); } }

Figure 11.3. Opening Windows

11.2.3. Closing Windows When the user closes an application-level window, the Client-Side Engine running in the browser will report the event to the server before the page is actually removed. You can catch the event with a Window.CloseListener, as is done in the example below. newWindow.addListener(new Window.CloseListener() { @Override public void windowClose(CloseEvent e) { // Do something. System.out.println(e.getWindow().getName() + " was closed"); // Add a text to the main window about closing. // (This does not update the main window.) getMainWindow().addComponent( new Label("Window '" + e.getWindow().getName() + "' was closed."));

Closing Windows

259


Advanced Web Application Topics

} });

Notice that the change to the server-side state of the main window (or another application-level window) does not refresh the window in the browser, so the change will be unseen until user interaction or polling refreshes the window. This problem and its dangers are discussed in Section 11.2.4, â&#x20AC;&#x153;Caveats in Using Multiple Windowsâ&#x20AC;? below. The close event does not occur if the browser crashes or the connection is otherwise severed violently. In such a situation, the window object will be left hanging, which could become a resource problem if you allow the users to open many such application-level windows. The positive side is that the user can reconnect to the window using the window URL.

11.2.4. Caveats in Using Multiple Windows Communication Between Windows For cases where you need communication between windows, we recommend using floating child windows. In Vaadin Release 5, an application window can not update the data in other windows. The contents of a window can only be updated when the particular window makes a request to the server. The request can be caused by user input or through polling. Changing the server-side state of a window while processing a user event from another window can potentially cause serious problems. Changing the client-side state of a window does not always immediately communicate the changes to the server. The server-side state can therefore be out of sync with the client-side state.

Figure 11.4. Communication Between Two Application-Level Windows

The following example creates a second window that changes the contents of the main window, as illustrated in the figure above. In this simple case, changing the main window contents is safe. // Create a table in the main window to hold items added // in the second window final Table table = new Table(); table.setPageLength(5); table.getSize().setWidth(100, Size.UNITS_PERCENTAGE); table.addContainerProperty("Name", String.class, ""); main.addComponent(table); // Create the second window final Window adderWindow = new Window("Add Items"); adderWindow.setName("win-adder"); main.getApplication().addWindow(adderWindow);

260

Caveats in Using Multiple Windows


Advanced Web Application Topics

// Create selection component to add items to the table final NativeSelect select = new NativeSelect("Select item to add"); select.setImmediate(true); adderWindow.addComponent(select); // Add some items to the selection String items[] = new String[]{"-- Select --", "Mercury", "Venus", "Earth", "Mars", "Jupiter", "Saturn", "Uranus", "Neptune"}; for (int i=0; i<items.length; i++) select.addItem(items[i]); select.setNullSelectionItemId(items[0]); // When an item is selected in the second window, add // table in the main window select.addListener(new ValueChangeListener() { public void valueChange(ValueChangeEvent event) { // If the selected value is something else // but a null selection item. if (select.getValue() != null) { // Add the selected item to the table // in the main window table.addItem(new Object[]{select.getValue()}, new Integer(table.size())); } } }); // Link to open the selection window Link link = new Link("Click to open second window", new ExternalResource(adderWindow.getURL()), "_new", 50, 200, Link.TARGET_BORDER_DEFAULT); main.addComponent(link); // Enable polling to update the main window ProgressIndicator poller = new ProgressIndicator(); poller.addStyleName("invisible"); main.addComponent(poller);

The example uses an invisible ProgressIndicator to implement polling. This is sort of a trick and a more proper API for polling is under design. Making the progress indicator invisible requires the following CSS style definition: .v-progressindicator-invisible { display: none; }

11.3. Embedding Applications in Web Pages Many web applications and especially web sites are not all AJAX, but AJAX is used only for specific functionalities. In practice, many web applications are a mixture of dynamic web pages and AJAX applications embedded to such pages. Embedding Vaadin applications is easy. There are two basic ways to embed them. One is to have a <div> placeholder for the web application and load the Vaadin Client-Side Engine with a simple JavaScript code.The second method is even easier, which is to simply use the <iframe> element. Both of these methods have advantages and disadvantages. The <div> method can only embed one application in a page, while the <iframe> method can embed as many as needed. One disadvantage of the <iframe> method is that the size of the <iframe> element is not flexible according to the content while the <div> method allows such flexibility. The following sections look closer into these two embedding methods.

Embedding Applications in Web Pages

261


Advanced Web Application Topics

11.3.1. Embedding Inside a div Element The loading code for the Client-Side Engine changed in IT Mill toolkit version 5.1.2 and the explanation below is no longer compatible with Vaadin. Please view the source code of the initial page of your application in your browser. You can embed a Vaadin application inside a web page with a method that is equivalent to loading the initial page content from the application servlet in a non-embedded application. Normally, the ApplicationServlet servlet generates an initial page that contains the correct parameters for the specific application. You can easily configure it to load multiple Vaadin applications on the same page, assuming that they use the same widget set. You can view the initial page for your application easily simply by opening the application in a web browser and viewing the HTML source code. You could just copy and paste the embedding code from the default initial page. It has, however, some extra functionality that is not normally needed: it generates some of the script content with document.write() calls, which is useful only when you are running the application as a portlet in a portal. The method outlined below is much simpler. The WebContent/multiapp.html file included in the Vaadin installation package provides an example of embedding (multiple) Vaadin applications in a page. After launching the demo application, you can view the example at URL http://localhost:8888/multiapp.html. Notice that the example assumes the use of root context for the applications (/). Embedding requires four elements inside the HTML document: 1. In the <head> element, you need to define the application URI and parameters and load the Vaadin Client-Side Engine. The vaadin variable is an associative map that can contain various runtime data used by the Client-Side Engine of Vaadin. The vaadinConfigurations item is itself an associate map that contains parameters for each of the applications embedded in the page. The map must contain the following items: appUri

The application URI consists of the context and the application path. If the context is /mycontext and the application path is myapp, the appUri would be /mycontext/myapp. The multiapp.html example assumes the use of root context, which is used in the demo application.

pathInfo

The PATHINFO parameter for the Servlet.

themeUri

URI of the application theme. The URI must include application context and the path to the theme directory. Themes are, by default, stored under the /VAADIN/themes/ path.

versionInfo

This item is itself an associative map that contains two parameters: vaadinVersion contains the version number of the Vaadin version used by the application. The applicationVersion parameter contains the version of the particular application.

The following example defines two applications to run in the same window: the Calculator and Hello World examples. In the example, the application context is /tk5. <script type="text/javascript"> var vaadin = { vaadinConfigurations: {

262

Embedding Inside a div Element


Advanced Web Application Topics

'calc': { appUri:'/tk5/Calc', pathInfo: '/', themeUri: '/tk5/VAADIN/themes/example', versionInfo : { vaadinVersion:"5.9.9-INTERNALNONVERSIONED-DEBUG-BUILD", applicationVersion:"NONVERSIONED" } }, 'hello': { appUri:'/tk5/HelloWorld', pathInfo: '/', themeUri: '/tk5/VAADIN/themes/example', versionInfo : { vaadinVersion:"5.9.9-INTERNALNONVERSIONED-DEBUG-BUILD", applicationVersion:"NONVERSIONED" } } }}; </script>

2. Loading the Vaadin Client-Side Engine is done with the following kind of line in the <head> element: <script language='javascript' src='/vaadin-examples/VAADIN/widgetsets/com.vaadin.terminal.gwt.DefaultWidgetSet/com.vaadin.terminal.gwt.DefaultWidgetSet.nocache.js'></script>

The engine URI consists of the context of the web application, vaadin-examples above, followed by the path to the JavaScript (.js) file of the widget set, relative to the WebContent directory. The file contains the Client-Side Engine compiled for the particular widget set. The line above assumes the use of the default widget set of Vaadin. If you have made custom widgets that are defined in a custom widget set, you need to use the path to the compiled widget set file. Widget sets must be compiled under the WebContent/VAADIN/widgetsets directory. 3. In the <html> element, you need to do a routine inclusion of GWT history iframe element as follows: <iframe id="__gwt_historyFrame" style="width:0;height:0;border:0"></iframe>

4. The location of the Vaadin application is defined with a div placeholder element having id="calc", where the identifier is the same as in the vaadinConfigurations parameter, as follows: <div id="calc"/>

Below is a complete example of embedding an application. It works out-of-the-box with the Calculator demo application. <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> <html xmlns="http://www.w3.org/1999/xhtml" > <head> <title>Embedding Example</title> <!-- Set parameters for the Vaadin Client-Side Engine. --> <script type="text/javascript"> var vaadin = {appUri:'Calc', pathInfo: '/'};

Embedding Inside a div Element

263


Advanced Web Application Topics

</script> <!-- Load the Vaadin Client-Side Engine. --> <script language='javascript' src='/vaadin-examples/VAADIN/widgetsets/com.vaadin.terminal.gwt.DefaultWidgetSet/com.vaadin.terminal.gwt.DefaultWidgetSet.nocache.js'></script>

<!-- We can stylize the web application. --> <style> #vaadin-ajax-window {background: #c0c0ff;} .v-button {background: pink;} </style> </head> <body> <!-- This <iframe> element is required by GWT. --> <iframe id="__gwt_historyFrame" style="width:0;height:0;border:0"></iframe> <h1>This is a HTML page</h1> <p>Below is the Vaadin application inside a table:</p> <table align="center" border="3" style="background: yellow;"> <tr><th>The Calculator</th></tr> <tr> <td> <!-- Placeholder <div> for the Vaadin application --> <div id="vaadin-ajax-window"/> </td> </tr> </table> </body> </html>

The page will look as shown in Figure 11.5, â&#x20AC;&#x153;Embedded Applicationâ&#x20AC;?. You can style the web application with themes as described in Chapter 8, Themes. The ClientSide Engine loads the style sheets required by the application. In addition, you can do styling in the embedding page, as was done in the example above. The Reservation Demo and Windowed Demos provide similar examples of embedding an application in a web page. The embedding web pages are WebContent/reservr.html and WebContent/windoweddemos.html, respectively. The disadvantage of this embedding method is that there can only be one web application embedded in a page. One is usually enough, but if it is not, you need to use the <iframe> method below.

11.3.2. Embedding Inside an iframe Element Embedding a Vaadin application inside an <iframe> element is even easier than the method described above, as it does not require definition of any Vaadin specific definitions. The use of <iframe> makes it possible to embed multiple web applications or two different views to the same application on the same page. You can embed an application with an element such as the following: <iframe src="/vaadin-examples/Calc"></iframe>

The problem with <iframe> elements is that their size of is not flexible depending on the content of the frame, but the content must be flexible to accommodate in the frame. You can set the size of an <iframe> element with height and width attributes.

264

Embedding Inside an iframe Element


Advanced Web Application Topics

Figure 11.5. Embedded Application

Below is a complete example of using the <iframe> to embed two applications in a web page. <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> <html xmlns="http://www.w3.org/1999/xhtml" > <head> <title>Embedding in IFrame</title> </head> <body style="background: #d0ffd0;"> <h1>This is a HTML page</h1> <p>Below are two Vaadin applications embedded inside a table:</p> <table align="center" border="3"> <tr> <th>The Calculator</th> <th>The Color Picker</th> </tr> <tr valign="top"> <td> <iframe src="/vaadin-examples/Calc" height="200" width="150" frameborder="0"></iframe> </td> <td> <iframe src="/vaadin-examples/colorpicker" height="330" width="400" frameborder="0"></iframe> </td> </tr> </table> </body> </html>

The page will look as shown in Figure 11.6, â&#x20AC;&#x153;Vaadin Applications Embedded Inside IFramesâ&#x20AC;? below.

Embedding Inside an iframe Element

265


Advanced Web Application Topics

Figure 11.6. Vaadin Applications Embedded Inside IFrames

11.4. Debug and Production Mode Vaadin applications can be run in two modes: debug mode and production mode. The debug mode, which is on by default, enables a number of built-in debug features for the developers. The features include: • Debug Window for accessing debug functionalities • Display debug information in the Debug Window and server console. • Analyze layouting button that analyzes the layout for possible problems. All applications are run in the debug mode by default (since IT Mill Toolkit version 5.3.0). The production mode can be enabled, and debug mode thereby disabled, by adding a productionMode=true parameter to the servlet context in the web.xml deployment descriptor: <context-param> <description>Vaadin production mode</description> <param-name>productionMode</param-name> <param-value>true</param-value> </context-param>

Enabling the production mode disables the debug features, thereby preventing users from easily inspecting the inner workings of the application from the browser.

11.4.1. Debug Mode Running an application in the debug mode enables the client-side Debug Window in the browser. You can open the Debug Window by adding "?debug" to the application URL, e.g.,

266

Debug and Production Mode


Advanced Web Application Topics

http://localhost:8080/myapp/?debug.The Debug Window, shown in Figure 11.7, â&#x20AC;&#x153;Debug Windowâ&#x20AC;?, consists of buttons controlling the debugging features and a scrollable log of debug messages.

Figure 11.7. Debug Window

Clear console

Clears the log in the Debug Window.

Restart app

Restarts the application.

Force layout

Causes all currently visible layouts to recalculate their appearance. Layout components calculate the space required by all child components, so the layout appearance must be recalculated whenever the size of a child component is changed. In normal applications, this is done automatically, but when you do themeing or alter the CSS with Firebug, you may need to force all layouts to recalculate themselves, taking into account the recently made changes.

Analyze layouts

This is described in the following section.

If you use the Firebug plugin in Mozilla Firefox, the log messages will also be printed to the Firebug console. In such a case, you may want to enable client-side debugging without showing the Debug Window with "?debug=quiet" in the URL. In the quiet debug mode, log messages will only be printed to the Firebug console.

11.4.2. Analyzing Layouts The Analyze layouts button analyzes the currently visible layouts and makes a report of possible layout related problems. All detected layout problems are displayed in the log and also printed to the console. The most common layout problem is caused by placing a component that has a relative size inside a container (layout) that has undefined size, e.g., adding a 100% wide Panel inside a HorizontalLayout with no width specification. In such a case, the error will look as shown below: Vaadin DEBUG - Window/1a8bd74 "My window" (width: MAIN WINDOW) - HorizontalLayout/1cf243b (width: UNDEFINED) - Panel/12e43f1 "My panel" (width: RELATIVE, 100.0 %) Layout problem detected: Component with relative width inside a HorizontalLayout with no width defined Relative sizes were replaced by undefined sizes, components may not render as expected.

Analyzing Layouts

267


Advanced Web Application Topics

This particular error tells that the Panel "My panel" is 100% wide while the width of the containing HorizontalLayout is undefined. The components will be rendered as if the the width of the contained Panel was undefined, which might not be what the developer wanted. There are two possible fixes for this case: if the Panel should fill the main window horizontally, set a width for the HorizontalLayout (e.g. 100% wide), or set the width of the Panel to "undefined" to render the it as it is currently rendered but avoiding the warning message. The same error is shown in the Debug Window in a slightly different form and with an additional feature (see Figure 11.8, “Debug Window Showing the Result of Analyze layouts.”). Checking the Emphasize component in UI box will turn red the background of the component that caused a warning, making it easy for the developer to figure out which component each warning relates to. The messages will also be displayed hierarchically, as a warning from a containing component often causes more warnings from its child components. A good rule of thumb is to work on the upper-level problems first and only after that worry about the warnings from the children.

Figure 11.8. Debug Window Showing the Result of Analyze layouts.

11.4.3. Custom Layouts CustomLayout components can not be analyzed in the same way as other layouts. For custom layouts, the Analyze layouts button analyzes all contained relative-sized components and checks if any relative dimension is calculated to zero so that the component will be invisible. The error log will display a warning for each of these invisible components. It would not be meaningful to emphasize the component itself as it is not visible, so when you select such an error, the parent layout of the component is emphasized if possible.

11.4.4. Debug Functions for Component Developers You can take advantage of the debug mode when developing client-side components. The static function ApplicationConnection.getConsole() will return a reference to a Console object which contains logging methods such as log(String msg) and error(String msg).These functions will print messages to the Debug Window and Firebug console in the same way as other debugging functionalities of Vaadin do. No messages will be printed if the Debug Window is not open or if the application is running in production mode.

11.5. Resources In addition to high-level resource classes described in Section 4.5, “Referencing Resources”, Vaadin provides low-level facilities for retrieving the URI and other parameters of HTTP requests. In the following, we will look into low-level interfaces for handling URIs and parameters to provide resources and functionalities.

268

Custom Layouts


Advanced Web Application Topics

Notice that using URI or parameter handlers to create "pages" is not meaningful in Vaadin or in AJAX applications generally. See Section 11.1, â&#x20AC;&#x153;Special Characteristics of AJAX Applicationsâ&#x20AC;? for reasons.

11.5.1. URI Handlers The URI parameter for the application is useful mainly for two purposes: for providing some special functionality according to the URI or for providing dynamic content. Dynamic content can also be provided with StreamResource. You can retrieve the URI for the HTTP request made for your application by implementing the com.vaadin.terminal.URIHandler interface. The handler class needs to be registered in the main window object of your application with the addURIHandler() method. You then get the URI by implementing the handleURI() method. The method gets two parameters: a context and a URI relative to the context. The context is the base URI for your application. public void init() { final Window main = new Window("Hello window"); setMainWindow(main); URIHandler uriHandler = new URIHandler() { public DownloadStream handleURI(URL context, String relativeUri) { // Do something here System.out.println("handleURI=" + relativeUri); // Should be null unless providing dynamic data. return null; } }; main.addURIHandler(uriHandler); }

If you have multiple URI handlers attached to a window, they are executed after one another. The URI handlers should return null, unless you wish to provide dynamic content with the call. Other URI handlers attached to the window will not be executed after some handler returns nonnull data. The combined parameter and URI handler example below shows how to create dynamic content with a URI handler. Notice that if you do provide dynamic content with a URI handler, the dynamic content is returned in the HTTP response. If the handler makes any changes to the UI state of the application, these changes are not rendered in the browser, as they are usually returned in the HTTP response made by the Application object and now the custom URI handler overrides the default behaviour. If your client-side code makes a server call that does update the UI state, the client-side must initiate an update from the server. For example, if you have an integration situation where you make a JavaScript call to the server, handle the request with a URI handler, and the server state changes as a side-effect, you can use the vaadin.forceSync() method to force the update.

11.5.2. Parameter Handlers You can retrieve the parameters passed to your application by implementing the com.vaadin.terminal.ParameterHandler interface.The handler class needs to be registered in the main window object of your application with the addParameterHandler() method. You then get the parameters in the handleParameters() method. The parameters are passes as a map from string key to a vector of string values.

URI Handlers

269


Advanced Web Application Topics

class MyParameterHandler implements ParameterHandler { public void handleParameters(Map parameters) { // Print out the parameters to standard output for (Iterator it = parameters.keySet().iterator(); it.hasNext();) { String key = (String) it.next(); String value = ((String[]) parameters.get(key))[0]; System.out.println("Key: "+key+", value: "+value); } } }

The parameter handler is not called if there are no parameters. Parameter handler is called before the URI handler, so if you handle both, you might typically want to just store the URI parameters in the parameter handler and do actual processing in URI handler. This allows you, for example, to create dynamic resources based on the URI parameters. import import import import import import import

java.awt.*; java.awt.image.BufferedImage; java.io.*; java.net.URL; java.util.Map; javax.imageio.ImageIO; com.vaadin.terminal.*;

/** * Demonstrates handling URI parameters and the URI itself to * create a dynamic resource. */ public class MyDynamicResource implements URIHandler, ParameterHandler { String textToDisplay = "- no text given -"; /** * Handle the URL parameters and store them for the URI * handler to use. */ public void handleParameters(Map parameters) { // Get and store the passed HTTP parameter. if (parameters.containsKey("text")) textToDisplay = ((String[])parameters.get("text"))[0]; } /** * Provides the dynamic resource if the URI matches the * resource URI. The matching URI is "/myresource" under * the application URI context. * * Returns null if the URI does not match. Otherwise * returns a download stream that contains the response * from the server. */ public DownloadStream handleURI(URL context, String relativeUri) { // Catch the given URI that identifies the resource, // otherwise let other URI handlers or the Application // to handle the response. if (!relativeUri.startsWith("myresource")) return null; // Create an image and draw some background on it. BufferedImage image = new BufferedImage (200, 200, BufferedImage.TYPE_INT_RGB); Graphics drawable = image.getGraphics(); drawable.setColor(Color.lightGray);

270

Parameter Handlers


Advanced Web Application Topics

drawable.fillRect(0,0,200,200); drawable.setColor(Color.yellow); drawable.fillOval(25,25,150,150); drawable.setColor(Color.blue); drawable.drawRect(0,0,199,199); // Use the parameter to create dynamic content. drawable.setColor(Color.black); drawable.drawString("Text: "+textToDisplay, 75, 100); try { // Write the image to a buffer. ByteArrayOutputStream imagebuffer = new ByteArrayOutputStream(); ImageIO.write(image, "png", imagebuffer); // Return a stream from the buffer. ByteArrayInputStream istream = new ByteArrayInputStream( imagebuffer.toByteArray()); return new DownloadStream (istream,null,null); } catch (IOException e) { return null; } } }

When you use the dynamic resource class in your application, you obviously need to provide the same instance of the class as both types of handler: MyDynamicResource myresource = new MyDynamicResource(); mainWindow.addParameterHandler(myresource); mainWindow.addURIHandler(myresource);

Figure 11.9. Dynamic Resource with URI Parameters

11.6. Shortcut Keys Vaadin provides simple ways for defining shortcut keys for field components and a default button, and a lower-level generic shortcut key binding API based on actions.

11.6.1. Click Shortcuts for Default Buttons You can add or set a click shortcut to a button to set it as "default" button; pressing the defined key, typically Enter, in any component in the window causes a click event for the button.

Shortcut Keys

271


Advanced Web Application Topics

You can define a click shortcut with the setClickShortcut() shorthand method: // Have an OK button and set it as the default button Button ok = new Button("OK"); ok.setClickShortcut(KeyCode.ENTER); ok.addStyleName("primary");

The primary style name highlights a button to show the default button status; usually with a bolder font than usual, depending on the theme. The result can be seen in Figure 11.10, “Default Button with Click Shortcut”.

Figure 11.10. Default Button with Click Shortcut

11.6.2. Field Focus Shortcuts You can define a shortcut key that sets the focus to a field component (any component that inherits AbstractField) by adding a FocusShortcut as a shortcut listener to the field. // A field with Alt+N bound to it TextField name = new TextField("Name (Alt+N)"); name.addShortcutListener( new AbstractField.FocusShortcut(name, KeyCode.N, ModifierKey.ALT)); layout.addComponent(name); // A field with Alt+A bound to it TextField address = new TextField("Address (Alt+A)"); address.addShortcutListener( new AbstractField.FocusShortcut(address, KeyCode.A, ModifierKey.ALT)); layout.addComponent(address);

The constructor of the FocusShortcut takes the field component as its first parameter, followed by the key code, and an optional list of modifier keys, as listed in Section 11.6.4, “Supported Key Codes and Modifier Keys”.

11.6.3. Generic Shortcut Actions Shortcut keys can be defined as actions using the ShortcutAction class. ShortcutAction extends the generic Action class that is used for example in Tree and Table for context menus. Currently, the only classes that accept ShortcutActions are Window and Panel. To handle key presses, you need to define an action handler by implementing the Handler interface. The interface has two methods that you need to implement: getActions() and handleAction(). The getActions() method must return an array of Action objects for the component, specified with the second parameter for the method, the sender of an action. For a keyboard shortcut, you use a ShortcutAction. The implementation of the method could be following: // Have the unmodified Enter key cause an event Action action_ok = new ShortcutAction("Default key", ShortcutAction.KeyCode.ENTER, null); // Have the C key modified with Alt cause an event Action action_cancel = new ShortcutAction("Alt+C",

272

Field Focus Shortcuts


Advanced Web Application Topics

ShortcutAction.KeyCode.C, new int[] { ShortcutAction.ModifierKey.ALT }); Action[] actions = new Action[] {action_cancel, action_ok}; public Action[] getActions(Object target, Object sender) { if (sender == myPanel) return actions; return null; }

The returned Action array may be static or you can create it dynamically for different senders according to your needs. The constructor of ShortcutAction takes a symbolic caption for the action; this is largely irrelevant for shortcut actions in their current implementation, but might be used later if implementors use them both in menus and as shortcut actions. The second parameter is the key code and the third a list of modifier keys, which are listed in Section 11.6.4, â&#x20AC;&#x153;Supported Key Codes and Modifier Keysâ&#x20AC;?. The following example demonstrates the definition of a default button for a user interface, as well as a normal shortcut key, Alt-C for clicking the Cancel button. public class DefaultButtonExample extends CustomComponent implements Handler { // Define and create user interface components Panel panel = new Panel("Login"); FormLayout formlayout = new FormLayout(); TextField username = new TextField("Username"); TextField password = new TextField("Password"); HorizontalLayout buttons = new HorizontalLayout(); // Create buttons and define their listener methods. Button ok = new Button("OK", this, "okHandler"); Button cancel = new Button("Cancel", this, "cancelHandler"); // Have the unmodified Enter key cause an event Action action_ok = new ShortcutAction("Default key", ShortcutAction.KeyCode.ENTER, null); // Have the C key modified with Alt cause an event Action action_cancel = new ShortcutAction("Alt+C", ShortcutAction.KeyCode.C, new int[] { ShortcutAction.ModifierKey.ALT }); public DefaultButtonExample() { // Set up the user interface setCompositionRoot(panel); panel.addComponent(formlayout); formlayout.addComponent(username); formlayout.addComponent(password); formlayout.addComponent(buttons); buttons.addComponent(ok); buttons.addComponent(cancel); // Set focus to username username.focus(); // Set this object as the action handler System.out.println("adding ah"); panel.addActionHandler(this); System.out.println("start done."); }

Generic Shortcut Actions

273


Advanced Web Application Topics

/** * Retrieve actions for a specific component. This method * will be called for each object that has a handler; in * this example just for login panel. The returned action * list might as well be static list. */ public Action[] getActions(Object target, Object sender) { System.out.println("getActions()"); return new Action[] { action_ok, action_cancel }; } /** * Handle actions received from keyboard. This simply directs * the actions to the same listener methods that are called * with ButtonClick events. */ public void handleAction(Action action, Object sender, Object target) { if (action == action_ok) { okHandler(); } if (action == action_cancel) { cancelHandler(); } } public void okHandler() { // Do something: report the click formlayout.addComponent(new Label("OK clicked. " + "User=" + username.getValue() + ", password=" + password.getValue())); } public void cancelHandler() { // Do something: report the click formlayout.addComponent(new Label("Cancel clicked. User=" + username.getValue() + ", password=" + password.getValue())); } }

Notice that the keyboard actions can currently be attached only to Panels and Windows. This can cause problems if you have components that require a certain key. For example, multi-line TextField requires the Enter key. There is currently no way to filter the shortcut actions out while the focus is inside some specific component, so you need to avoid such conflicts.

11.6.4. Supported Key Codes and Modifier Keys The shortcut key definitions require a key code to identify the pressed key and modifier keys, such as Shift, Alt, or Ctrl, to specify a key combination. The key codes are defined in the ShortcutAction.KeyCode interface and are:

274

Keys A to Z

Normal letter keys

F1 to F12

Function keys

BACKSPACE, DELETE, ENTER, ESCAPE, INSERT, TAB

Control keys

NUM0 to NUM9

Number pad keys

Supported Key Codes and Modifier Keys


Advanced Web Application Topics

ARROW_DOWN, ARROW_UP, ARROW_LEFT, ARROW_RIGHT

Arrow keys

HOME, END, PAGE_UP, PAGE_DOWN

Other movement keys

Modifier keys are defined in ShortcutAction.ModifierKey and are: ModifierKey.ALT

Alt key

ModifierKey.CTRL

Ctrl key

ModifierKey.SHIFT

Shift key

All constructors and methods accepting modifier keys take them as a variable argument list following the key code, separated with commas. For example, the following defines a Ctrl-Shift-N key combination for a shortcut. TextField name = new TextField("Name (Ctrl+Shift+N)"); name.addShortcutListener( new AbstractField.FocusShortcut(name, KeyCode.N, ModifierKey.CTRL, ModifierKey.SHIFT));

Supported Key Combinations The actual possible key combinations vary greatly between browsers, as most browsers have a number of built-in shortcut keys, which can not be used in web applications. For example, Mozilla Firefox allows binding almost any key combination, while Opera does not even allow binding Alt shortcuts. Other browsers are generally in between these two. Also, the operating system can reserve some key combinations and some computer manufacturers define their own system key combinations.

11.7. Printing Vaadin does not currently have any special support for printing. Printing on the server-side is, in any case, largely independent from the web UI of an application. You just have to take care that the printing does not block server requests, possibly by running printing in another thread. For client-side printing, most browsers support printing the web page. Vaadin does not explicitly support launching the printing in browser, but you can easily use the JavaScript print() method that opens the print window of the browser. final Button print = new Button("Print This Page"); print.addListener(new ClickListener() { public void buttonClick(ClickEvent event) { print.getWindow().executeJavaScript("print();"); } });

This button would print the current page, including the button itself. Often, you want to be able to print a report or receipt and it should not have any visible UI components. In such a case, you could offer it as a PDF resource, or you could open a new window, as is done below, and automatically launch printing. // A button to open the printer-friendly page. Button print = new Button("Click to Print"); print.addListener(new Button.ClickListener() {

Printing

275


Advanced Web Application Topics

public void buttonClick(ClickEvent event) { // Create a window that contains what you want to print Window window = new Window("Window to Print"); // Have some content to print window.addComponent(new Label( "<h1>Here's some dynamic content</h1>\n" + "<p>This is to be printed to the printer.</p>", Label.CONTENT_XHTML)); // Add the printing window as a new application-level // window getApplication().addWindow(window); // Open it as a popup window with no decorations getWindow().open(new ExternalResource(window.getURL()), "_blank", 500, 200, // Width and height Window.BORDER_NONE); // No decorations // Print automatically when the window opens. // This call will block until the print dialog exits! window.executeJavaScript("print();"); // Close the window automatically after printing window.executeJavaScript("self.close();"); } });

How the browser opens the window, as an actual (popup) window or just a tab, depends on the browser. Notice that calling the print() method in the window will block the entire application until the print dialog exits. After printing, we automatically close the window with another JavaScript call, as there is no close() method in Window. Printing as PDF would not require creating a Window object, but you would need to provide the content as a static or a dynamic resource for the open() method. Printing a PDF file would obviously require a PDF viewer cabability (such as Adobe Reader) in the browser.

11.8. Portal Integration Vaadin supports running applications as portlets, as defined in the JSR-168 (Java Portlet API) and JSR-286 (Java Portlet API 2.0) standards. While providing generic support for all portals implementing the standards, Vaadin especially supports the Liferay portal and the needed portalspecific configuration is given below for Liferay. You can deploy the Vaadin demo package WAR (available from the download site) directly to a portal such as Liferay. It contains all the necessary portlet configuration files. For optimal performance with Liferay, you can install the Vaadin library and other needed resources in Liferay as described later in this section. You can find more documentation and examples from the Vaadin Developer's Site at http://dev.vaadin.com/.

11.8.1. Deploying to a Portal Deploying a Vaadin application as a portlet is essentially just as easy as deploying a regular application to an application server. You do not need to make any changes to the application itself, but only the following: â&#x20AC;˘ Application packaged as a WAR

276

Portal Integration


Advanced Web Application Topics

• • • •

• WEB-INF/portlet.xml descriptor • WEB-INF/web.xml descriptor for Portlet 1.0 portlets • WEB-INF/liferay-portlet.xml descriptor for Liferay • WEB-INF/liferay-display.xml descriptor for Liferay • WEB-INF/liferay-plugin-package.properties for Liferay Widget set installed to portal (optional) Themes installed to portal (optional) Vaadin library installed to portal (optional) Portal configuration settings (optional)

Installing the widget set and themes to the portal is required for running two or more Vaadin portlets simultaneously in a single portal page. As this situation occurs quite easily, we recommend installing them in any case. In addition to the Vaadin library, you will need to copy the portlet.jar to your project. It is included in the Vaadin installation package. Notice that you must not put the portlet.jar in the same WebContent/WEB-INF/lib directory as the Vaadin JAR or otherwise include it in the WAR to be deployed, because it would create a conflict with the internal portlet library of the portal. How you actually deploy a WAR package depends on the portal. In Liferay, you simply drop it to the deploy subdirectory under the Liferay installation directory. The deployment depends on the application server under which Liferay runs; for example, if you use Liferay bundled with Tomcat, you will find the extracted package in the webapps directory under the Tomcat installation directory included in Liferay.

11.8.2. Creating a Portal Application Project in Eclipse While you can create the needed deployment descriptors manually for any existing Vaadin application, as described in subsequent sections, the Vaadin Plugin for Eclipse provides a wizard for easy creation of portal application projects. Creation of a portal application project is almost identical to the creation of a regular application project. For a full treatment of the New Project Wizard and the possible options, please see Section 2.4.1, “Creating the Project”. 1. Start creating a new project by selecting from the menu File → New → Project....

Creating a Portal Application Project in Eclipse

277


Advanced Web Application Topics 2. In the New Project window that opens, select Web â&#x2020;&#x2019; Vaadin Project and click Next. 3. In the Vaadin Project step, you need to set the basic web project settings. You need to give at least the project name, the runtime, and select Generic Portlet for the deployment configuration; the default values should be good for the other settings.

You can click Finish here to use the defaults for the rest of the settings, or click Next. 4. The settings in the Web Module step define the basic servlet-related settings and the structure of the web application project. All the settings are pre-filled, and you should normally accept them as they are and click Next. 5. The Vaadin project step page has various Vaadin-specific application settings. These are largely the same as for regular applications.You should not need to change anything as you can change the application titles and other details afterwards. The Create portlet template option should be automatically selected. You can give another portlet title of you want. You can change most of the settings afterward.

278

Creating a Portal Application Project in Eclipse


Advanced Web Application Topics

Create project template

Creates an application class and all the needed portlet deployment descriptors.

Application name

The application name is used in the title of the main window (which is usually invisible in portlets) and as an identifier, either as is or with a suffix, in various deployment descriptors.

Base package name

Java package for the application class.

Application class name

Name of the application class. The default is derived from the project name.

Portlet version

Same as in the project settings.

Portlet title

The portlet title, defined in portlet.xml, can be used as the display name of the portlet (at least in Liferay). The default value is the project name. The title is also used as a short description in liferay-plugin-package.properties.

Vaadin version

Same as in the project settings.

Finally, click Finish to create the project. 6. Eclipse may ask you to switch to J2EE perspective. A Dynamic Web Project uses an external web server and the J2EE perspective provides tools to control the server and manage application deployment. Click Yes.

11.8.3. Portlet Deployment Descriptors To deploy a portlet WAR in a portal, you need to provide the basic portlet.xml descriptor specified in the Java Portlet standard. In addition, you may need to include possible portal vendor specific deployment descriptors. The ones required by Liferay are described below.

Portlet 2.0 Deployment Descriptor The por tlet WAR must include a por tlet descr iptor located at WebContent/WEB-INF/portlet.xml. A portlet definition includes the portlet name, mapping to a servlet in web.xml, modes supported by the portlet, and other configuration. Below is an example of a simple portlet definition in portlet.xml descriptor. <?xml version="1.0" encoding="UTF-8" standalone="no"?> <portlet-app xmlns="http://java.sun.com/xml/ns/portlet/portlet-app_2_0.xsd" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" version="2.0" xsi:schemaLocation= "http://java.sun.com/xml/ns/portlet/portlet-app_2_0.xsd http://java.sun.com/xml/ns/portlet/portlet-app_2_0.xsd"> <portlet> <portlet-name>Portlet Example portlet</portlet-name> <display-name>Vaadin Portlet Example</display-name> <!-- Map portlet to a servlet. --> <portlet-class> com.vaadin.terminal.gwt.server.ApplicationPortlet2 </portlet-class>

Portlet Deployment Descriptors

279


Advanced Web Application Topics

<init-param> <name>application</name> <!-- The application class with package name. --> <value>com.example.myportlet.MyportletApplication</value> </init-param> <!-- Supported portlet modes and content types. --> <supports> <mime-type>text/html</mime-type> <portlet-mode>view</portlet-mode> <portlet-mode>edit</portlet-mode> <portlet-mode>help</portlet-mode> </supports> <!-- Not always required but Liferay requires these. --> <portlet-info> <title>Vaadin Portlet Example</title> <short-title>Portlet Example</short-title> </portlet-info> </portlet> </portlet-app>

Listing supported portlet modes in portlet.xml enables the corresponding portlet controls in the portal user interface that allow changing the mode, as described later.

Portlet 1.0 Deployment Descriptor The portlet deployment descriptor for Portlet 1.0 API is largely the same as for Portlet 2.0. The main differences are: 1. XML namespace and schema names 2. Portlet-class: ApplicationPortlet vs ApplicationPortlet2 3. The application parameter is a name of the servlet (defined in web.xml in Portlet 1.0, but name of the application class in Portlet 2.0. There is no longer a separate web.xml file in Servlet 2.0. 4. The portlet-name must not be same as the servlet name in Portlet 1.0; in Portlet 2.0 this does not matter. 5. Below is an example of a complete deployment descriptor for Portlet 1.0: <?xml version="1.0" encoding="UTF-8"?> <portlet-app version="1.0" xmlns="http://java.sun.com/xml/ns/portlet/portlet-app_1_0.xsd" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation= "http://java.sun.com/xml/ns/portlet/portlet-app_1_0.xsd http://java.sun.com/xml/ns/portlet/portlet-app_1_0.xsd"> <portlet> <!-- Must not be the same as servlet name. --> <portlet-name>Portlet Example portlet</portlet-name> <display-name>Vaadin Portlet Example</display-name> <!-- Map portlet to a servlet. --> <portlet-class> com.vaadin.terminal.gwt.server.ApplicationPortlet

280

Portlet Deployment Descriptors


Advanced Web Application Topics

</portlet-class> <init-param> <name>application</name> <!-- Must match the servlet URL mapping in web.xml. --> <value>portletexample</value> </init-param> <!-- Supported portlet modes and content types. --> <supports> <mime-type>text/html</mime-type> <portlet-mode>view</portlet-mode> <portlet-mode>edit</portlet-mode> <portlet-mode>help</portlet-mode> </supports> <!-- Not always required but Liferay requires these. --> <portlet-info> <title>Vaadin Portlet Example</title> <short-title>Portlet Example</short-title> </portlet-info> </portlet> </portlet-app>

The value of the application parameter must match the context in the <url-pattern> element in the <servlet-mapping> in the web.xml deployment descriptor, without the path qualifiers in the pattern. The above example would match the following servlet mapping in web.xml: <servlet-mapping> <servlet-name>Portlet Example</servlet-name> <url-pattern>/portletexample/*</url-pattern> </servlet-mapping>

In fact, it would also match the /* mapping.

Using a Single Widget Set If you have just one Vaadin application that you ever need to run in your portal, you can just deploy the WAR as described above and that's it. However, if you have multiple applications, especially ones that use different custom widget sets, you run into problems, because a portal window can load only a single Vaadin widget set at a time. You can solve this problem by combining all the different widget sets in your different applications into a single widget set using inheritance or composition. For example, the portal demos defined in the portlet.xml in the demo WAR have the following setting for all portlets so that they will all use the same widget set: <portlet> ... <!-- Use the portal default widget set for all portal demos. --> <init-param> <name>widgetset</name> <value>com.vaadin.portal.gwt.PortalDefaultWidgetSet</value> </init-param> ...

The PortalDefaultWidgetSet extends SamplerWidgetSet, which extends the DefaultWidgetSet. The DefaultWidgetSet is therefore essentially a subset of PortalDefaultWidgetSet, which contains also the widgets required by the Sampler demo. Other applications that would otherwise require only the regular DefaultWidgetSet, and do not define their own widgets, can just as well use the larger set, making them compatible with the demos. The PortalDefaultWidgetSet will also be the default Vaadin widgetset bundled in Liferay 5.3 and later.

Portlet Deployment Descriptors

281


Advanced Web Application Topics

If your portlets are contained in multiple WARs, which can happen quite typically, you need to install the widget set and theme portal-wide so that all the portlets can use them. See Section 11.8.5, “Installing Vaadin in Liferay” on configuring the widget sets in the portal itself.

Liferay Portlet Descriptor Liferay requires a special liferay-portlet.xml descriptor file that defines Liferay-specific parameters. Especially, Vaadin portlets must be defined as "instanceable", but not "ajaxable". Below is an example descriptor for the earlier portlet example: <?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE liferay-portlet-app PUBLIC "-//Liferay//DTD Portlet Application 4.3.0//EN" "http://www.liferay.com/dtd/liferay-portlet-app_4_3_0.dtd"> <liferay-portlet-app> <portlet> <!-- Matches definition in portlet.xml. --> <!-- Note: Must not be the same as servlet name. --> <portlet-name>Portlet Example portlet</portlet-name> <instanceable>true</instanceable> <ajaxable>false</ajaxable> </portlet> </liferay-portlet-app>

See Liferay documentation for further details on the liferay-portlet.xml deployment descriptor.

Liferay Display Descriptor The WEB-INF/liferay-display.xml file defines the portlet category under which portlets are located in the Add Application window in Liferay. Without this definition, portlets will be organized under the "Undefined" category. The following display configuration, which is included in the demo WAR, puts the Vaadin portlets under the "Vaadin" category, as shown in Figure 11.11, “Portlet Categories in Add Application Window”. <?xml version="1.0"?> <!DOCTYPE display PUBLIC "-//Liferay//DTD Display 4.0.0//EN" "http://www.liferay.com/dtd/liferay-display_4_0_0.dtd"> <display> <category name="Vaadin"> <portlet id="Portlet Example portlet" /> </category> </display>

282

Portlet Deployment Descriptors


Advanced Web Application Topics

Figure 11.11. Portlet Categories in Add Application Window

See Liferay documentation for further details on how to configure the categories in the liferay-display.xml deployment descriptor.

Liferay Plugin Package Properties The liferay-plugin-package.properties file defines a number of settings for the portlet, most importantly the Vaadin JAR to be used. name=Portlet Example portlet short-description=myportlet module-group-id=Vaadin module-incremental-version=1 #change-log= #page-uri= #author= license=Proprietary portal-dependency-jars=\ vaadin.jar

name

The plugin name must match the portlet name.

short-description

A short description of the plugin. This is by default the project name.

module-group-id

The application group, same as the category id defined in liferay-display.xml.

license

The plugin license type; "proprietary" by default.

portal-dependency-jars

The JAR libraries on which this portlet depends. This should have value vaadin.jar, unless you need to use a specific version. The JAR must be installed in the portal,

Portlet Deployment Descriptors

283


Advanced Web Application Topics

for example, in Liferay bundled with Tomcat to tomcat-x.x.x/webapps/ROOT/WEB-INF/lib/vaadin.jar.

11.8.4. Portlet Hello World The Hello World program that runs as a portlet is no different from a regular Vaadin application, as long as it doesn't need to handle portlet actions, mode changes, and so on. import com.vaadin.Application; import com.vaadin.ui.*; public class PortletExample extends Application { @Override public void init() { Window mainWindow = new Window("Portlet Example"); Label label = new Label("Hello Vaadin user"); mainWindow.addComponent(label); setMainWindow(mainWindow); } }

In addition to the application class, you need the descriptor files, libraries, and other files as described earlier. Figure 11.12, “Portlet Project Structure in Eclipse” shows the complete project structure under Eclipse. Installed as a portlet in Liferay from the Add Application menu, the application will show as illustrated in Figure 11.13, “Hello World Portlet”.

Figure 11.13. Hello World Portlet

11.8.5. Installing Vaadin in Liferay Loading widget sets, themes, and the Vaadin JAR from a portlet is possible as long as you have a single portlet, but causes a problem if you have multiple portlets. To solve this, Vaadin portlets need to use a globally installed widget set, themes, and Vaadin JAR. They, and all the required configuration, are bundled with Liferay 5.3 and later, but if you are using an earlier version of Liferay or use a custom widget set, custom themes, or a specific version of Vaadin, you will need to do the configuration manually. In these instructions, we assume that you use Liferay bundled with Apache Tomcat, although you can use many other application servers with Liferay just as well. The Tomcat installation is included in the Liferay installation package, under the tomcat-x.x.x directory. The Vaadin JAR should be put in tomcat-x.x.x/webapps/ROOT/WEB-INF/lib/vaadin.jar. The Vaadin version number should normally be left out from the JAR.

284

Portlet Hello World


Advanced Web Application Topics

Figure 11.12. Portlet Project Structure in Eclipse

The widget set needs to be located at /html/VAADIN/widgetsets/ and themes at /html/VAADIN/themes/ path under the portal context. You simply need to copy the contents from under your WebContent/VAADIN directory to the tomcat-x.x.x/webapps/ROOT/html/VAADIN directory under the Liferay installation directory. If you use a built-in widget set or theme included in Vaadin, such as the PortalDefaultWidgetSet, you should copy it from the Vaadin installation directory, from under WebContent/VAADIN/widgetsets. The default themes are located under WebContent/VAADIN/themes in the installation directory. You need to define the widget set, the theme, and the JAR in the portal-ext.properties configuration file for Liferay, as described earlier. The file should normally be placed in the Liferay installation directory. See Liferay documentation for details on the configuration file. Below is an example of a portal-ext.properties file: # Path under which the VAADIN directory is located. # (/html is the default so it is not needed.) # vaadin.resources.path=/html # Portal-wide widget set vaadin.widgetset=com.vaadin.portal.gwt.PortalDefaultWidgetSet # Theme to use vaadin.theme=reindeer

The allowed parameters are:

Installing Vaadin in Liferay

285


Advanced Web Application Topics

vaadin.resources.path

Specifies the resource root path under the portal context. This is /html by default. Its actual location depends on the portal and the application server; in Liferay with Tomcat it would be located at webapps/ROOT/html under the Tomcat installation directory.

vaadin.widgetset

The widget set class to use. Give the full path to the class name in the dot notation. If the parameter is not given, the default widget set is used.

vaadin.theme

Name of the theme to use. If the parameter is not given, the default theme is used, which is reindeer in Vaadin 6.

You will need to restart Liferay after creating or modifying the portal-ext.properties file.

11.8.6. Handling Portlet Requests Portals such as Liferay are not AJAX applications but reload the page every time a user interaction requires data from the server. They consider a Vaadin application to be a regular web application that works by HTTP requests. All the AJAX communications required by the Vaadin application are done by the Vaadin Client-Side Engine (the widget set) past the portal, so that the portal is unaware of the communications. The only way a portal can interact with an application is to load it with a HTTP request; reloading does not reset the application. The Portlet 2.0 API supports four types of requests: render, action, resource, and event requests.The old Portlet 1.0 API supports only the render and action requests. Requests can be caused by user interaction with the portal controls or by clicking action URLs displayed by the portlet. You can handle portlet requests by implementing the PortletListener interface and the handler methods for each of the request types. You can use the request object passed to the handler to access certain portal data, such as user information, the portlet mode, etc. The PortletListener interface is defined in the PortletApplicationContext2 for Portlet 2.0 API and com.vaadin.terminal.gwt.server.PortletApplicationContext class for the old Portlet 1.0 API. You can get the portlet application context with getContext() method of the application class. You need to have the portlet.jar in your class path during development. However, you must not deploy the portlet.jar with the portlet, because it would create a conflict with the internal portlet library of the portal. You should put it in a directory that is not deployed with the portlet, for example, if you are using Eclipse, under the lib directory under the project root, not under WebContent/WEB-INF/lib, for example. You can also define portal actions that you can handle in the handleActionRequest() method of the interface. You add your portlet request listener to the application context of your application, which is a PortletApplicationContext when (and only when) the application is being run as a portlet. // Check that we are running as a portlet. if (getContext() instanceof PortletApplicationContext2) { PortletApplicationContext2 ctx = (PortletApplicationContext2) getContext(); // Add a custom listener to handle action and // render requests. ctx.addPortletListener(this, new MyPortletListener());

286

Handling Portlet Requests


Advanced Web Application Topics

} else { getMainWindow().showNotification( "Not initialized via Portal!", Notification.TYPE_ERROR_MESSAGE); }

The handler methods receive references to request and response objects, which are defined in the Java Servlet API. Please refer to the Servlet API documentation for further details. The PortletDemo application included in the demo WAR package includes examples of processing mode and portlet window state changes in a portlet request listener.

11.8.7. Handling Portlet Mode Changes Portals support three portlet modes defined in the Portlet API: view, edit, and help modes. The view mode is the default and the portal can have buttons to switch the portlet to the other modes. In addition to the three predefined modes, the Portlet API standards allow custom portlet modes, although portals may support custom modes to a varying degree. You need to define which portlet modes are enabled in the portlet.xml deployment descriptor as follows. <!-- Supported portlet modes and content types. --> <supports> <mime-type>text/html</mime-type> <portlet-mode>view</portlet-mode> <portlet-mode>edit</portlet-mode> <portlet-mode>help</portlet-mode> </supports>

Changes in the portlet mode are received as resource requests, which you can handle with a handleResourceRequest(), defined in the PortletListener interface. The current portlet mode can be acquired with getPortletMode() from the request object. The following complete example (for Portlet 2.0) shows how to handle the three built-modes in a portlet application. // Use Portlet 2.0 API import com.vaadin.terminal.gwt.server.PortletApplicationContext2; import com.vaadin.terminal.gwt.server.PortletApplicationContext2.PortletListener; public class PortletModeExample extends Application implements PortletListener { Window mainWindow; ObjectProperty data; // Data to view and edit VerticalLayout viewContent = new VerticalLayout(); VerticalLayout editContent = new VerticalLayout(); VerticalLayout helpContent = new VerticalLayout(); @Override public void init() { mainWindow = new Window("Myportlet Application"); setMainWindow(mainWindow); // Data model data = new ObjectProperty("<h1>Heading</h1>"+ "<p>Some example content</p>"); // Prepare views for the three modes (view, edit, help) // Prepare View mode content Label viewText = new Label(data, Label.CONTENT_XHTML); viewContent.addComponent(viewText);

Handling Portlet Mode Changes

287


Advanced Web Application Topics

// Prepare Edit mode content RichTextArea editText = new RichTextArea(); editText.setCaption("Edit the value:"); editText.setPropertyDataSource(data); editContent.addComponent(editText); // Prepare Help mode content Label helpText = new Label("<h1>Help</h1>" + "<p>This helps you!</p>", Label.CONTENT_XHTML); helpContent.addComponent(helpText); // Start in the view mode mainWindow.setContent(viewContent); // Check that we are running as a portlet. if (getContext() instanceof PortletApplicationContext2) { PortletApplicationContext2 ctx = (PortletApplicationContext2) getContext(); // Add a custom listener to handle action and // render requests. ctx.addPortletListener(this, this); } else { mainWindow.showNotification("Not running in portal", Notification.TYPE_ERROR_MESSAGE); } } // Dummy implementations for the irrelevant request types public void handleActionRequest(ActionRequest request, ActionResponse response, Window window) { } public void handleRenderRequest(RenderRequest request, RenderResponse response, Window window) { } public void handleEventRequest(EventRequest request, EventResponse response, Window window) { } public void handleResourceRequest(ResourceRequest request, ResourceResponse response, Window window) { // Switch the view according to the portlet mode if (request.getPortletMode() == PortletMode.EDIT) window.setContent(editContent); else if (request.getPortletMode() == PortletMode.VIEW) window.setContent(viewContent); else if (request.getPortletMode() == PortletMode.HELP) window.setContent(helpContent); } }

Figure 11.14, “Portlet Modes in Action” shows the resulting portlet in the three modes: view, edit, and help. In Liferay, the edit mode is shown in the popup menu as a Preferences item.

288

Handling Portlet Mode Changes


Advanced Web Application Topics

Figure 11.14. Portlet Modes in Action

11.8.8. Non-Vaadin Portlet Modes In some cases, it can be useful to implement certain modes of a portlet as pure HTML or JSP pages instead of running the full Vaadin application user interface in them. Common reasons for this are static pages (e.g. a simple help mode), integrating legacy content to a portlet (e.g. a JSP configuration interface) and providing an ultra-lightweight initial view for a portlet (for users behind slow connections). Fully static modes that do not require the Vaadin server side application to be running can be implemented by subclassing the portlet class ApplicationPortlet2 (Portlet 2.0). The subclass can either create the HTML content directly or dispatch the request to e.g. a HTML or JSP page via the portal. When using this approach, any Vaadin portlet and portlet request listeners are not called. Customizing the content for the standard modes (view, edit, and help) can be performed by overriding the methods doView, doEdit and doHelp, respectively. Custom modes can be handled by implementing similar methods with the @javax.portlet.RenderMode(name = "mymode") annotation.

Non-Vaadin Portlet Modes

289


Advanced Web Application Topics

You need to define which portlet modes are enabled in the portlet.xml deployment descriptor as described in Section 11.8.7, â&#x20AC;&#x153;Handling Portlet Mode Changesâ&#x20AC;?. Also, the portlet class in portlet.xml should point to the customized subclass of ApplicationPortlet2. The following example (for Portlet 2.0) shows how to create a static help page for the portlet. portlet.xml: <!-- Supported portlet modes and content types. --> <supports> <mime-type>text/html</mime-type> <portlet-mode>view</portlet-mode> <portlet-mode>help</portlet-mode> </supports>

HtmlHelpPortlet.java:: // Use Portlet 2.0 API import com.vaadin.terminal.gwt.server.ApplicationPortlet2; public class HtmlHelpPortlet extends ApplicationPortlet2 { // override the help mode, let the Vaadin application handle the view mode @Override protected void doHelp(RenderRequest request, RenderResponse response) throws PortletException, IOException { // bypass the Vaadin application entirely response.setContentType("text/html"); response.getWriter().println("This is the help text as plain HTML."); // // // //

alternatively could use the dispatcher for e.g. JSP help pages: PortletRequestDispatcher dispatcher = getPortletContext() .getRequestDispatcher("/html/myhelp.jsp"); dispatcher.include(request, response);

} }

To produce pure HTML portlet content from a running Vaadin application instead of statically outside an application, the ApplicationPortlet2 method writeAjaxPage should be overridden. This approach allows using the application state in HTML content generation, and all relevant Vaadin portlet request and portlet listeners are called around the portlet content generation. However, the client side engine (widgetset) is not loaded by the browser, which can shorten the initial page display time. <portlet-class>com.vaadin.demo.portlet.HtmlModePortlet</portlet-class> <supports> <mime-type>text/html</mime-type> <portlet-mode>view</portlet-mode> <portlet-mode>help</portlet-mode> </supports> public class CountApplication extends Application { private int count = 0; public void init() { Window window = new Window("Portlet mode example"); window.addComponent(new Label("This is the Vaadin application.")); window.addComponent(new Label("Try opening the help mode.")); setMainWindow(window); } public int incrementCount() { return ++count;

290

Non-Vaadin Portlet Modes


Advanced Web Application Topics

} } // Use Portlet 2.0 API public class HtmlModePortlet extends AbstractApplicationPortlet { @Override protected void writeAjaxPage(RenderRequest request, RenderResponse response, Window window, Application application) throws PortletException, IOException { if (PortletMode.HELP.equals(request.getPortletMode())) { CountApplication app = (CountApplication) application; response.setContentType("text/html"); response.getWriter().println("This is the HTML help, shown " + app.incrementCount() + " times so far."); } else { super.writeAjaxPage(request, response, window, application); } } @Override protected Class<? extends Application> getApplicationClass() { return CountApplication.class; } }

The user can freely move between Vaadin and non-Vaadin portlet modes with the user interface provided by the portal (for standard modes) or the portlet (e.g. action links). Once the server side application has been started, it continues to run as long as the session is alive. If necessary, specific portlet mode transitions can be disallowed in portlet.xml. In the case of Portlet 1.0, both a portlet and a servlet are involved. A render request is received by ApplicationPortlet when the portlet mode is changed, and serving pure HTML in some modes can be achieved by overriding the method render and handling the modes of interest separately while calling super.render() for other modes. As always, when extending the portlet, the reference to the portlet class in portlet.xml needs to be updated. To serve HTML-only content in the Portlet 1.0 case after starting the server side application and calling the relevant listeners, the servlet class ApplicationServlet should be subclassed instead of the portlet. The method writeAjaxPage can be overridden to produce custom HTML content for certain modes. However, it should be noted that some HTML content (e.g. loading the portalwide Vaadin theme) is created by the portlet and not the servlet.

11.9. Google App Engine Integration Vaadin includes support to run Vaadin applications in the Google App Engine (GAE). The most essential requirement for GAE is the ability to serialize the application state. Vaadin applications are serializable through the java.io.Serializable interface. To run as a GAE application, an application must use GAEApplicationServlet instead of ApplicationServlet in web.xml, and of course implement the java.io.Serializable interface for all persistent classes. You also need to enable session support in appengine-web.xml with: <sessions-enabled>true</sessions-enabled>

The Vaadin Project wizard can create the configuration files needed for GAE deployment. See Section 2.4.1, â&#x20AC;&#x153;Creating the Projectâ&#x20AC;?. When the Google App Engine deployment configuration is selected, the wizard will create the project structure following the GAE Servlet convention instead of the regular Servlet convention. The main differences are:

Google App Engine Integration

291


Advanced Web Application Topics

• Source directory: src/main/java • Output directory: war/WEB-INF/classes • Content directory: war

Rules and Limitations Running Vaadin applications in Google App Engine has the following rules and limitations: • Avoid using the session for storage, usual App Engine limitations apply (no synchronization, that is, it is unreliable). • Vaadin uses memcache for mutex, the key is of the form _vmutex<sessionid>. • The Vaadin WebApplicationContext class is serialized separately into memcache and datastore; the memcache key is _vac<sessionid> and the datastore entity kind is _vac with identifiers of the type _vac<sessionid>. • Do not update the application state when serving an ApplicationResource (such as ClassResource.getStream()). • Avoid (or be very careful when) updating application state in a TransactionListener it is called even when the application is not locked and won't be serialized (such as with ApplicationResource), and changes can therefore be lost (it should be safe to update things that can be safely discarded later, that is, valid only for the current request). • The application remains locked during uploads - a progress bar is not possible.

11.10. Common Security Issues 11.10.1. Sanitizing User Input to Prevent Cross-Site Scripting You can put raw XHTML content in many components, such as the Label and CustomLayout, as well as in tooltips and notifications. In such cases, you should make sure that if the content has any possibility to come from user input, the input is well sanitized before displaying it. Otherwise, a malicious user can easily make a cross-site scripting attack by injecting offensive JavaScript code in such components. Offensive code can easily be injected with <script> markup or in tag attributes as events, such as onLoad. Cross-site scripting vulnerabilities are browser dependent, depending on the situations in which different browsers execute scripting markup. There is no generic way to sanitize user input as different applications can allow different kinds of input. Pruning (X)HTML tags out is somewhat simple, but some applications may need to allow (X)HTML. It is therefore the responsibility of the application to sanitize the input. Character encoding can make sanitization more difficult, as offensive tags can be encoded so that they are not recognized by a sanitizer. This can be done, for example, with HTML character entities and with variable-width encodings such as UTF-8 or various CJK encodings, by abusing multiple representations of a character. Most trivially, you could input < and > with &lt; and &gt;, respectively. The input could also be malformed and the sanitizer must be able to interpret it exactly as the browser would, and different browsers can interpret malformed HTML and variablewidth character encodings differently.

292

Rules and Limitations


Advanced Web Application Topics

Notice that the problem applies also to user input from a RichTextArea is transmitted as XHTML from the browser to server-side and is not sanitized. As the entire purpose of the RichTextArea component is to allow input of formatted text, you can not just remove all HTML tags. Also many attributes, such as style, should pass through the sanitization.

11.11. URI Fragment and History Management with UriFragmentUtility A major issue in AJAX applications is that as they run in a single web page, bookmarking the application URL (or more generally the URI) can only bookmark the application, not an application state. This is a problem for many applications such as product catalogs and forums, in which it would be good to provide links to specific products or messages. Consequently, as browsers remember the browsing history by URI, the history and the Back button do not normally work. The solution is to use the fragment part of the URI, which is separated from the primary part (address + path + optional query parameters) of the URI with the hash (#) character. For example: http://example.com/path#myfragment

The exact syntax of the fragment part is defined in RFC 3986 (Internet standard STD 66) that defines the URI syntax. A fragment may only contain the regular URI path characters (see the standard) and additionally the slash and the question mark. The UriFragmentUtility is a special-purpose component that manages the URI fragment; it allows setting the fragment and to handle user-made changes to it. As it is a regular component, though invisible, you must add it to a layout in an application window with the addComponent(), as usual. public void init() { Window main = new Window("URI Fragment Example"); setMainWindow(main); // Create the URI fragment utility final UriFragmentUtility urifu = new UriFragmentUtility(); main.addComponent(urifu);

Notice that the utility component can work only when it is attached to the window, so in practice it must be added in the init() method of the application and must afterwards always remain in the application's user interface. You can set the URI fragment with the setFragment() method of the UriFragmentUtility object. The method takes the fragment as a string parameter. In the following example, we have a menu, from which the user can select the URI fragment. // Application state menu final ListSelect menu = new ListSelect("Select a URI Fragment"); menu.addItem("mercury"); menu.addItem("venus"); menu.addItem("earth"); menu.addItem("mars"); menu.setImmediate(true); main.addComponent(menu); // Set the URI Fragment when menu selection changes menu.addListener(new Property.ValueChangeListener() { public void valueChange(ValueChangeEvent event) { String itemid = (String) event.getProperty().getValue(); urifu.setFragment(itemid); } });

URI Fragment and History Management with UriFragmentUtility

293


Advanced Web Application Topics

The URI fragment and any changes to it are passed to an application as FragmentChangedEvents, which you can handle with a FragmentChangedListener. You can get the new fragment value with the getFragment() method from the URI fragment utility component. // When the URI fragment is given, use it to set menu selection urifu.addListener(new FragmentChangedListener() { public void fragmentChanged(FragmentChangedEvent source) { String fragment = source.getUriFragmentUtility().getFragment(); if (fragment != null) menu.setValue(fragment); } });

Figure 11.15, â&#x20AC;&#x153;Application State Management with URI Fragment Utilityâ&#x20AC;? shows an application that allows specifying the menu selection with a URI fragment and correspondingly sets the fragment when the user selects a menu item, as done in the code examples above.

Figure 11.15. Application State Management with URI Fragment Utility

11.12. Capturing HTTP Requests Behind the event-driven processing model of Vaadin lies the Java Servlet API, which is based on processing HTTP requests. These requests are normally hidden from Vaadin applications, but can be caught using the HttpServletRequestListener interface. You must implement the interface in your application class.The two methods defined in the interface, onRequestStart() and onRequestEnd(), allow processing the request before and after other processing. import javax.servlet.http.HttpServletRequest; import javax.servlet.http.HttpServletResponse; import com.vaadin.Application; import com.vaadin.terminal.gwt.server.HttpServletRequestListener; import com.vaadin.ui.*; public class HttpServletRequestApplication extends Application implements HttpServletRequestListener { @Override public void init() { System.out.println("

Application.init() called.");

Window main = new Window("URI Fragment Example"); setMainWindow(main); setTheme("book-examples"); // Does nothing but causes a request Button button = new Button ("Make a request"); main.addComponent(button);

294

Capturing HTTP Requests


Advanced Web Application Topics

} public void onRequestStart(HttpServletRequest request, HttpServletResponse response) { System.out.println("[Start of request"); System.out.println(" Query string: " + request.getQueryString()); System.out.println(" Path: " + request.getPathInfo()); } public void onRequestEnd(HttpServletRequest request, HttpServletResponse response) { System.out.println(" End of request]"); } }

The onRequestStart() is called for the first time when the application class is loaded but the init() is not yet called. This can be seen in the output of the above code example: [Start of request Query string: null Path: null Application.init() called. End of request] [Start of request Query string: repaintAll=1&sh=1050&sw=1680&cw=500&ch=300&vw=500 Path: /UIDL/ End of request] [Start of request Query string: windowName=1071684214 Path: /UIDL/ End of request]

The first call is a regular HTML page load, so the URL path is simply the application path. The subsequent calls are AJAX calls made using the UIDL protocol, so the request path includes the /UIDL/ part. This is important to know when using cookies, as explained later.

11.12.1. Using Request and Response Objects The HttpServletRequest object provides access to the request data, such as request headers, path info, and query string, as well as to some higher-level information such as cookies. The HttpServletResponse object is somewhat different, as most write operations write data directly to the output stream of the server request. It is therefore possible to add new headers and cookies in the onRequestStart(), and make other settings, but not later on, especially not in the onRequestEnd(), as all the UIDL response data has already been written to the output stream. The framework writes the UIDL response to the output stream of the response before calling onRequestEnd(). You therefore have to be careful when writing to the response object. You can usually write to it when handling component events in listeners, as is done in the cookie example later. While it is theoretically possible to redirect the output stream of the response object to write custom data to the response, you should never need to do that, as it would break the UIDL communication protocol. The servlet request and response objects are defined in the Java Servlet API. Please refer to its documentation for more detailed information.

Using Request and Response Objects

295


Advanced Web Application Topics

11.12.2. Managing Cookies Setting and reading cookies is one of the typical uses of HttpServletRequestListener. The application gets the HttpServletRequest object containing the cookies in the onRequestStart() method.

Setting a Cookie You normally set a cookie in an event listener. As the request object is a transient object that exists only for the duration of the request, it is not accessible from the Application object. The only way to access it is to store it in onRequestStart(), as done in the following example. public class CookieExampleApplication extends Application implements HttpServletRequestListener { HttpServletResponse response; public void onRequestStart(HttpServletRequest request, HttpServletResponse response) { // Store the reference to the response object for // using it in event listeners this.response = response; ... } ...

We can then use the reference to set or delete cookies in event listeners. Notice that the cookie path property is automatically set to the application path (such as /book-examples/cookies) on the first request, but contains the UIDL subpath on subsequent calls (such as /book-examples/cookies/UIDL). As the cookies are matched against this path, you may need to set the path explicitly with setPath(). newuser = new TextField ("Give a user name"); login = new Button("Login"); login.addListener(new Button.ClickListener() { public void buttonClick(ClickEvent event) { Object value = newuser.getValue(); if (value != null && !((String)value).equals("")) { username = (String) value; Cookie cookie = new Cookie("username", username); // Use a fixed path cookie.setPath("/book-examples"); cookie.setMaxAge(3600); // One hour response.addCookie(cookie); System.out.println("Set cookie."); newuser.setEnabled(false); login.setEnabled(false); restart.setEnabled(true); logout.setEnabled(true); } } }); loginrow.addComponent(newuser); loginrow.addComponent(login);

Removing cookie can be set in similar way by setting the maxAge property to zero. // Delete the cookie Cookie cookie = new Cookie("username", username); cookie.setPath("/book-examples");

296

Managing Cookies


Advanced Web Application Topics

cookie.setMaxAge(0); // Delete response.addCookie(cookie);

Reading a Cookie Reading a cookie can be done in the onRequestStart() event. As this method is called also on the first client request before the application is initialized, it is possible to read user identification cookies and such on the first request. public class CookieExampleApplication extends Application implements HttpServletRequestListener { String username; public void onRequestStart(HttpServletRequest request, HttpServletResponse response) { if (username == null) { Cookie[] cookies = request.getCookies(); for (int i=0; i<cookies.length; i++) { if (cookies[i].getName().equals("username")) // Log the user in automatically username = cookies[i].getValue(); } } } ...

Notice that the request path is the application path (such as /book-examples/cookies) on the first request, but contains the UIDL subpath on subsequent AJAX calls (such as /book-examples/cookies/UIDL). So, if you have set the cookie in an AJAX request without setting the cookie path explicitly to such that does not contain the UIDL subpath, the cookie will be filtered out on the initial onRequestStart() call.

11.13. Drag and Drop Dragging an object from one location to another by grabbing it with mouse, holding the mouse button pressed, and then releasing the button to "drop" it to the other location is a common way to move, copy, or associate objects. For example, most operating systems allow dragging and dropping files between folders or dragging a document on a program to open it. In Vaadin, it is possible to drag and drop components and parts of certain components. Dragged objects, or transferables, are essentially data objects. You can drag and drop rows in Table and nodes in Tree components, either within or between the components. You can also drag entire components by wrapping them inside DragAndDropWrapper. Dragging starts from a drag source, which defines the transferable. Transferables implement the Transferable interfaces. For trees and tables, which are bound to Container data sources, a node or row transferable is a reference to an Item in the Vaadin Data Model. Dragged components are referenced with a WrapperTransferable. Starting dragging does not require any clientserver communication, you only need to enable dragging. All drag and drop logic occurs in two operations: determining (accepting) where dropping is allowed and actually dropping. Drops can be done on a drop target, which implements the DropTarget interface. Three components implement the interface: Tree, Table, and DragAndDropWrapper. These accept and drop operations need to be provided in a drop handler. Essentially all you need to do to enable drag and drop is to enable dragging in the drag source and implement the getAcceptCriterion() and drop() methods in the DropHandler interface. The client-server architecture of Vaadin causes special requirements for the drag and drop functionality. The logic for determining where a dragged object can be dropped, that is, accepting

Drag and Drop

297


Advanced Web Application Topics

a drop, should normally be done on the client-side, in the browser. Server communications are too slow to have much of such logic on the server-side. The drag and drop feature therefore offers a number of ways to avoid the server communications to ensure a good user experience.

11.13.1. Handling Drops Most of the user-defined drag and drop logic occurs in a drop handler, which is provided by implementing the drop() method in the DropHandler interface. A closely related definition is the drop accept criterion, which is defined in the getAcceptCriterion() method in the same interface. It is described in Section 11.13.4, â&#x20AC;&#x153;Accepting Dropsâ&#x20AC;? later. The drop() method gets a DragAndDropEvent as its parameters. The event object provides references to two important object: Transferable and TargetDetails. A Transferable contains a reference to the object (component or data item) that is being dragged. A tree or table item is represented as a TreeTransferable or TableTransferable object, which carries the item identifier of the dragged tree or table item. These special transferables, which are bound to some data in a container, are DataBoundTransferable. Dragged components are represented as WrapperTransferable objects, as the components are wrapped in a DragAndDropWrapper. The TargetDetails object provides details regarding the exact location where the transferable object is being dropped. If the target is a tree or a table, the TreeTargetDetails and TableTargetDetails objects provide the tree or table item on which the drop is being made. For entire components, the information is provided in a WrapperDropDetails object. In addition to the target item or component, the details objects provide a drop location. For selection components, the location can be obtained with the getDropLocation() and for wrapped components with verticalDropLocation() and horizontalDropLocation(). The locations are specified as either VerticalDropLocation or HorizontalDropLocation objects. The drop location objects specify whether the transferable is being dropped above, below, or directly on (at the middle of) a component or item. Dropping on a Tree, Table, and a wrapped component is explained further in the following sections.

11.13.2. Dropping Items On a Tree You can drag items from, to, or within a Tree. Making tree a drag source requires simply setting the drag mode with setDragMode(). Tree currently supports only one drag mode, TreeDragMode.NODE, which allows dragging single tree nodes. While dragging, the dragged node is referenced with a TreeTransferable object, which is a DataBoundTransferable. The tree node is identified by the item ID of the container item. When a transferable is dropped on a tree, the drop location is stored in a TreeTargetDetails object, which identifies the target location by item ID of the tree node on which the drop is made. You can get the item ID with getItemIdOver() method in AbstractSelectTargetDetails, which the TreeTargetDetails inherits. A drop can occur directly on or above or below a node; the exact location is a VerticalDropLocation, which you can get with the getDropLocation() method. In the example below, we have a Tree and we allow reordering the tree items by drag and drop. final Tree tree = new Tree("Inventory"); tree.setContainerDataSource(TreeExample.createTreeContent()); layout.addComponent(tree); // Expand all items for (Iterator<?> it = tree.rootItemIds().iterator(); it.hasNext();)

298

Handling Drops


Advanced Web Application Topics

tree.expandItemsRecursively(it.next()); // Set the tree in drag source mode tree.setDragMode(TreeDragMode.NODE); // Allow the tree to receive drag drops and handle them tree.setDropHandler(new DropHandler() { public AcceptCriterion getAcceptCriterion() { return AcceptAll.get(); } public void drop(DragAndDropEvent event) { // Wrapper for the object that is dragged Transferable t = event.getTransferable(); // Make sure the drag source is the same tree if (t.getSourceComponent() != tree) return; TreeTargetDetails target = (TreeTargetDetails) event.getTargetDetails(); // Get ids of the dragged item and the target item Object sourceItemId = t.getData("itemId"); Object targetItemId = target.getItemIdOver(); // On which side of the target the item was dropped VerticalDropLocation location = target.getDropLocation(); HierarchicalContainer container = (HierarchicalContainer) tree.getContainerDataSource(); // Drop right on an item -> make it a child if (location == VerticalDropLocation.MIDDLE) tree.setParent(sourceItemId, targetItemId); // Drop at the top of a subtree -> make it previous else if (location == VerticalDropLocation.TOP) { Object parentId = container.getParent(targetItemId); container.setParent(sourceItemId, parentId); container.moveAfterSibling(sourceItemId, targetItemId); container.moveAfterSibling(targetItemId, sourceItemId); } // Drop below another item -> make it next else if (location == VerticalDropLocation.BOTTOM) { Object parentId = container.getParent(targetItemId); container.setParent(sourceItemId, parentId); container.moveAfterSibling(sourceItemId, targetItemId); } } });

Accept Criteria for Trees Tree defines some specialized accept criteria for trees. TargetInSubtree (client-side)

Accepts if the target item is in the specified sub-tree. The sub-tree is specified by the item ID of the root of the sub-tree in the constructor. The second constructor includes a depth parameter, which specifies how deep from the given root node are drops accepted. Value -1 means infinite, that is, the entire sub-tree, and is therefore the same as the simpler constructor.

Dropping Items On a Tree

299


Advanced Web Application Topics

TargetItemAllowsChildren (client-side)

Accepts a drop if the tree has setChildrenAllowed() enabled for the target item. The criterion does not require parameters, so the class is a singleton and can be acquired with Tree.TargetItemAllowsChildren.get(). For example, the following composite criterion accepts drops only on nodes that allow children, but between all nodes: return new Or (Tree.TargetItemAllowsChildren.get(), new Not(VerticalLocationIs.MIDDLE));

TreeDropCriterion (server-side) Accepts drops on only some items, which as specified by a set of item IDs.You must extend the abstract class and implement the getAllowedItemIds() to return the set. While the criterion is server-side, it is lazyloading, so that the list of accepted target nodes is loaded only once from the server for each drag operation. See Section 11.13.4, “Accepting Drops” for an example. In addition, the accept criteria defined in AbstractSelect are available for a Tree, as listed in Section 11.13.4, “Accepting Drops”.

11.13.3. Dropping Items On a Table You can drag items from, to, or within a Table. Making table a drag source requires simply setting the drag mode with setDragMode(). Table supports dragging both single rows, with TableDragMode.ROW, and multiple rows, with TableDragMode.MULTIROW. While dragging, the dragged node or nodes are referenced with a TreeTransferable object, which is a DataBoundTransferable. Tree nodes are identified by the item IDs of the container items. When a transferable is dropped on a table, the drop location is stored in a AbstractSelectTargetDetails object, which identifies the target row by its item ID. You can get the item ID with getItemIdOver() method. A drop can occur directly on or above or below a row; the exact location is a VerticalDropLocation, which you can get with the getDropLocation() method from the details object.

Accept Criteria for Tables Table defines one specialized accept criterion for tables. TableDropCriterion (server-side) Accepts drops only on (or above or below) items that are specified by a set of item IDs. You must extend the a b s t ra c t class and implement the getAllowedItemIds() to return the set. While the criterion is server-side, it is lazy-loading, so that the list of accepted target items is loaded only once from the server for each drag operation.

11.13.4. Accepting Drops You can not drop the objects you are dragging around just anywhere. Before a drop is possible, the specific drop location on which the mouse hovers must be accepted. Hovering a dragged

300

Dropping Items On a Table


Advanced Web Application Topics

object over an accepted location displays an accept indicator, which allows the user to position the drop properly. As such checks have to be done all the time when the mouse pointer moves around the drop targets, it is not feasible to send the accept requests to the server-side, so drops on a target are normally accepted by a client-side accept criterion. A drop handler must define the criterion on the objects which it accepts to be dropped on the target.The criterion needs to be provided in the getAcceptCriterion() method of the DropHandler interface. A criterion is represented in an AcceptCriterion object, which can be a composite of multiple criteria that are evaluated using logical operations. There are two basic types of criteria: client-side and server-side criteria. The various built-in criteria allow accepting drops based on the identity of the source and target components, and on the data flavor of the dragged objects. To allow dropping any transferable objects, you can return a universal accept criterion, which you can get with AcceptAll.get(). tree.setDropHandler(new DropHandler() { public AcceptCriterion getAcceptCriterion() { return AcceptAll.get(); } ...

Client-Side Criteria The client-side criteria, which inherit the ClientSideCriterion, are verified on the client-side, so server requests are not needed for verifying whether each component on which the mouse pointer hovers would accept a certain object. The following client-side criteria are define in com.vaadin.event.dd.acceptcriterion: AcceptAll

Accepts all transferables and targets.

And

Logical AND operation on two client-side criterion; accepts the transferable if all the defined sub-criteria accept it.

ContainsDataFlavour

The transferable must contain the defined data flavour.

Not

Logical NOT operation on two client-side criterion; accepts the transferable if and only if the sub-criterion does not accept it.

Or

Logical OR operation on two client-side criterion; accepts the transferable if any of the defined sub-criteria accept it.

SourceIs

Accepts all transferables from any of the given source components

SourceIsTarget

Accepts the transferable only if the source component is the same as the target. This criterion is useful for ensuring that items are dragged only within a tree or a table, and not from outside it.

TargetDetailIs

Accepts any transferable if the target detail, such as the item of a tree node or table row, is of the given data flavor and has the given value.

In addition, target components such as Tree and Table define some component-specific clientside accept criteria. See Section 11.13.2, â&#x20AC;&#x153;Dropping Items On a Treeâ&#x20AC;? for more details.

Accepting Drops

301


Advanced Web Application Topics

AbstractSelect defines the following criteria for all selection components, including Tree and Table. AcceptItem

Accepts only specific items from a specific selection component. The selection component, which must inherit AbstractSelect, is given as the first parameter for the constructor. It is followed by a list of allowed item identifiers in the drag source.

AcceptItem.ALL

Accepts all transferables as long as they are items.

TargetItemIs

Accepts all drops on the specified target items. The constructor requires the target component (AbstractSelect) followed by a list of allowed item identifiers.

VerticalLocationIs.MIDDLE, TOP, and BOTTOM

The three static criteria accepts drops on, above, or below an item. For example, you could accept drops only in between items with the following: public AcceptCriterion getAcceptCriterion() { return new Not(VerticalLocationIs.MIDDLE); }

Server-Side Criteria The server-side criteria are verified on the server-side with the accept() method of the ServerSideCriterion class. This allows fully programmable logic for accepting drops, but the negative side is that it causes a very large amount of server requests. A request is made for every target position on which the pointer hovers. This problem is eased in many cases by the component-specific lazy loading criteria TableDropCriterion and TreeDropCriterion. They do the server visit once for each drag and drop operation and return all accepted rows or nodes for current Transferable at once. The accept() method gets the drag event as a parameter so it can perform its logic much like in drop(). public AcceptCriterion getAcceptCriterion() { // Server-side accept criterion that allows drops on any other // location except on nodes that may not have children ServerSideCriterion criterion = new ServerSideCriterion() { public boolean accept(DragAndDropEvent dragEvent) { TreeTargetDetails target = (TreeTargetDetails) dragEvent.getTargetDetails(); // The tree item on which the load hovers Object targetItemId = target.getItemIdOver(); // On which side of the target the item is hovered VerticalDropLocation location = target.getDropLocation(); if (location == VerticalDropLocation.MIDDLE) if (! tree.areChildrenAllowed(targetItemId)) return false; // Not accepted return true; // Accept everything else } }; return criterion; }

302

Accepting Drops


Advanced Web Application Topics

The server-side criteria base class ServerSideCriterion provides a generic accept() method. The more specific TableDropCriterion and TreeDropCriterion are conveniency extensions that allow definiting allowed drop targets as a set of items. They also provide some optimization by lazy loading, which reduces server communications significantly. public AcceptCriterion getAcceptCriterion() { // Server-side accept criterion that allows drops on any // other tree node except on node that may not have children TreeDropCriterion criterion = new TreeDropCriterion() { @Override protected Set<Object> getAllowedItemIds( DragAndDropEvent dragEvent, Tree tree) { HashSet<Object> allowed = new HashSet<Object>(); for (Iterator<Object> i = tree.getItemIds().iterator(); i.hasNext();) { Object itemId = i.next(); if (tree.hasChildren(itemId)) allowed.add(itemId); } return allowed; } }; return criterion; }

Accept Indicators When a dragged object hovers on a drop target, an accept indicator is displayed to show whether or not the location is accepted. For MIDDLE location, the indicator is a box around the target (tree node, table row, or component). For vertical drop locations, the accepted locations are shown as horizontal lines, and for horizontal drop locations as vertical lines. You can disable the accept indicators or drag hints with the no-vertical-drag-hints, no-horizontal-drag-hints, and no-box-drag-hints styles, as shown in the following example. wrapper.addStyleName("no-vertical-drag-hints"); wrapper.addStyleName("no-horizontal-drag-hints"); wrapper.addStyleName("no-box-drag-hints");

11.13.5. Dragging Components Dragging a component requires wrapping the source component within a DragAndDropWrapper. You can then allow dragging by putting the wrapper (and the component) in drag mode with setDragStartMode(). The method supports two drag modes: DragStartMode.WRAPPER and DragStartMode.COMPONENT, which defines whether the entire wrapper is shown as the drag image while dragging or just the wrapped component. // Have a component to drag final Button button = new Button("An Absolute Button"); // Put the component in a D&D wrapper and allow dragging it final DragAndDropWrapper buttonWrap = new DragAndDropWrapper(button); buttonWrap.setDragStartMode(DragStartMode.COMPONENT); // Set the wrapper to wrap tightly around the component buttonWrap.setSizeUndefined(); // Add the wrapper, not the component, to the layout layout.addComponent(buttonWrap, "left: 50px; top: 50px;");

Dragging Components

303


Advanced Web Application Topics

The default height of DragAndDropWrapper is undefined, but the default width is 100%. If you want to ensure that the wrapper fits tightly around the wrapped component, you should call setSizeUndefined() for the wrapper. Doing so, you should make sure that the wrapped component does not have a relative size, which would cause a paradox. Dragged components are referenced in the WrapperTransferable. You can get the reference to the dragged component with getDraggedComponent(). The method will return null if the transferable is not a component. Also HTML 5 drags (see later) are held in wrapper transferables.

11.13.6. Dropping on a Component Drops on a component are enabled by wrapping the component in a DragAndDropWrapper. The wrapper is an ordinary component; the constructor takes the wrapped component as a parameter. You just need to define the DropHandler for the wrapper with setDropHandler(). In the following example, we allow moving components in an absolute layout. Details on the drop handler are given later. // A layout that allows moving its contained components // by dragging and dropping them final AbsoluteLayout absLayout = new AbsoluteLayout(); absLayout.setWidth("100%"); absLayout.setHeight("400px"); ... put some (wrapped) components in the layout ... // Wrap the layout to allow handling drops DragAndDropWrapper layoutWrapper = new DragAndDropWrapper(absLayout); // Handle moving components within the AbsoluteLayout layoutWrapper.setDropHandler(new DropHandler() { public AcceptCriterion getAcceptCriterion() { return AcceptAll.get(); } public void drop(DragAndDropEvent event) { ... } });

Target Details for Wrapped Components The drop handler receives the drop target details in a WrapperTargetDetails object, which implements the TargetDetails interface. public void drop(DragAndDropEvent event) { WrapperTransferable t = (WrapperTransferable) event.getTransferable(); WrapperTargetDetails details = (WrapperTargetDetails) event.getTargetDetails();

The wrapper target details include a MouseEventDetails object, which you can get with getMouseEvent(). You can use it to get the mouse coordinates for the position where the mouse button was released and the drag ended. Similarly, you can find out the drag start position from the transferable object (if it is a WrapperTransferable) with getMouseDownEvent(). // Calculate the drag coordinate difference int xChange = details.getMouseEvent().getClientX() - t.getMouseDownEvent().getClientX(); int yChange = details.getMouseEvent().getClientY()

304

Dropping on a Component


Advanced Web Application Topics

- t.getMouseDownEvent().getClientY(); // Move the component in the absolute layout ComponentPosition pos = absLayout.getPosition(t.getSourceComponent()); pos.setLeftValue(pos.getLeftValue() + xChange); pos.setTopValue(pos.getTopValue() + yChange);

You can get the absolute x and y coordinates of the target wrapper with getAbsoluteLeft() and getAbsoluteTop(), which allows you to translate the absolute mouse coordinates to coordinates relative to the wrapper. Notice that the coordinates are really the position of the wrapper, not the wrapped component; the wrapper reserves some space for the accept indicators. The verticalDropLocation() and horizontalDropLocation() return the more detailed drop location in the target.

11.13.7. Dragging Files from Outside the Browser The DragAndDropWrapper allows dragging files from outside the browser and dropping them on a component wrapped in the wrapper. Dropped files are automatically uploaded to the application and can be acquired from the wrapper with getFiles(). The files are represented as Html5File objects as defined in the inner class. You can define an upload Receiver to receive the content of a file to an OutputStream. Dragging and dropping files to browser is supported in HTML 5 and requires a compatible browser, such as Mozilla Firefox 3.6 or newer.

11.14. Using Add-on Components In addition to the built-in components, layouts, themes, and data sources, many others are available as add-ons, either from the Vaadin Directory or from independent sources. Both commercial and free components exist. Installation of themes, data sources, and components built with server-side component composition is simple, just dropping a JAR package in a project and, usually, compiling the included widget set (the client-side implementation).

11.14.1. Downloading Add-ons from Vaadin Directory Vaadin Directory at http://vaadin.com/directory/ provides a rich collection of add-ons for Vaadin. You can download Directory add-on packages from the details page of an add-on. 1. Select the version; some add-ons have several versions available. The latest is shown by default, but you can choose another the version to download from the dropdown menu in the header of the details page. 2. Click Download Now and save the JAR or Zip file on your computer. 3. If the add-on is packaged in a Zip package, unzip the package and follow any instructions provided inside the package. Typically, you just need to copy a JAR file to your web project under the WEB-INF/lib directory. 4. Update and recompile your project. In Eclipse, select the project and press F5. 5. You need to compile the client-side implementations of the add-on components, that is, a widget set. This is the case for majority of add-ons, except for pure server-side,

Dragging Files from Outside the Browser

305


Advanced Web Application Topics

theme, or data binding add-ons. You must recompile the widget set if you install a new version of the add-on or the Vaadin library. See the subsequent sections for detailed instructions for compiling widget sets. 6. Update the project in web server and possibly restart the server. After trying out an add-on, you can give some feedback to the author of the add-on by rating the add-on with one to five stars and optionally leaving a comment. Please note the add-on license. While most commercial add-ons can be downloaded directly, you should note their license and other terms and conditions. Many are offered under a dual licensing agreement so that they can be used in open source projects for free, and many have a trial period for closed-source development.

11.14.2. Compiling Add-on Widget Sets in Eclipse To be able to compile widget sets in Eclipse, you need to have the Vaadin Plugin for Eclipse installed, as instructed in Section 2.2.5, “Vaadin Plugin for Eclipse”. An application can only have one widget set, so if you use multiple add-ons and possibly your own custom widgets, they need to be combined to a single widget set that inherits them. You can create the combining widget set manually and the Eclipse plugin simply update it when you add new add-ons. Otherwise, the Eclipse plugin automatically creates a project-specific widget set under the project root source folder. The Eclipse plugin compiles widget sets automatically by default. They do not normally need to be recompiled after changes to server-side classes, so if the automatic recompilation get annoying, disable it from the project settings. To compile the widget set(s) manually, click the Compile Vaadin widgets button in Eclipse toolbar or press Ctrl-6. You must recompile the widget set(s) always when you install a new version of the add-on or of the Vaadin library.

Figure 11.16. The Compile Vaadin widgets Button in Eclipse Toolbar

Further information on defining and compiling widget sets is given in Section 10.2.2, “Recompiling the Widget Set”, Section 10.5, “Defining a Widget Set”, and Section 10.8.4, “Compiling GWT Widget Sets”.

11.14.3. Compiling Add-on Widget Sets with an Ant Script The Vaadin installation package (the Zip package) includes an example Ant build script for compiling a widget set. The script is located in WebContent/docs/example-source/build-widgetset.xml. Once you have unpacked the installation package and changed to the directory in a command-line window, you can enter: $ ant -f WebContent/docs/example-source/build-widgetset.xml

You can copy the build script to your project. See Section 10.8.4, “Compiling GWT Widget Sets” for details on configuring the build script and the available build targets, and Section 10.5, “Defining a Widget Set” for information on the widget set definition file.

306

Compiling Add-on Widget Sets in Eclipse


Advanced Web Application Topics

If you are using an IDE such as Eclipse, always remember to refresh the project after compiling the widget set.

11.14.4. Troubleshooting If you experience problems, do the following: • Check the .gwt.xml widget set definition file under the widget set folder in the project root package. For example, if the project root package is com.example.myproject, the widget set definition file would be com.example.myproject.widgetset.MyprojectWidgetset.gwt.xml. See Section 10.5, “Defining a Widget Set” for details on the contents of the widget set definition file. • Check the WEB-INF/web.xml deployment descriptor and see that the servlet for your application has a widget set parameter, such as the following: <init-param> <description>Application widgetset</description> <param-name>widgetset</param-name> <param-value>com.example.myproject.widgetset.MyprojectWidgetset</param-value> </init-param>

• See the VAADIN/widgetsets directory and check that the widget set appears there. You can remove it and recompile to see that compilation works properly. • Use the Net tab in Firebug to see that the widget set (and theme) is loaded properly. • Use the ?debug parameter for the application to see if there is any version conflict between the widget set and the Vaadin library, or the themes. See Section 11.4.1, “Debug Mode” for details. • Refresh and recompile the project. In Eclipse, select the project and press F5, stop the server, clean the server temporary directories, and restart it. • Check the Error Log view in Eclipse (or the IDE you use). For more specific problems related to widget sets definition and compilation, see Section 10.8.8, “Troubleshooting”.

11.14.5. Removing Widget Sets Version mismatch problems with custom widget sets are a common source of grief for many beginners in Vaadin. If you need add-ons or your own custom components that include widget sets, you of course need to compile them, but otherwise it is unnecessary. If you do not use any such add-ons or your own custom components, do the following: 1. If you are using the Eclipse plugin, disable automatic widget set compilation from project preferences, in the Vaadin category, by selecting Suspend automatic widgetset builds. This prevents accidental compilation of the unnecessary widget sets. You may want to do this anyhow as the automatic builds can be annoying. You can still always build the widget set with the button in the toolbar. 2. Remove all widget set folders from under the VAADIN/widgetsets folder.

Troubleshooting

307


Advanced Web Application Topics

3. Edit the WEB-INF/web.xml file and remove the widgetset init parameter from the servlet. It looks as follows: <init-param> <description>Application widgetset</description> <param-name>widgetset</param-name> <param-value>com.vaadin.demo.colorpicker.widgetset.ColorPickerWidgetSet</param-value> </init-param>

4. Refresh the project. In Eclipse, select the project and press F5, stop the server, clean the server temporary directories, and restart it. At least in development environments, if you have extracted Vaadin themes to the VAADIN/themes folder, you should remove them and let them be loaded dynamically from the Vaadin JAR.

308

Removing Widget Sets


Appendix A

User Interface Definition Language (UIDL) User Interface Definition Language (UIDL) is a language for serializing user interface contents and changes in responses from web server to a browser. The idea is that the server-side components "paint" themselves to the screen (a web page) with the language. The UIDL messages are parsed in the browser and translated to GWT widgets. The UIDL is used through both server-side and client-side APIs. The server-side API consists of the PaintTarget interface, described in Section A.1, “API for Painting Components”. The clientside interface depends on the implementation of the client-side engine. In Vaadin Release 5, the client-side engine uses the Google Web Toolkit framework. Painting the user interface with a GWT widget is described in Section 10.3, “Google Web Toolkit Widgets”. UIDL supports painting either the entire user interface or just fragments of it. When the application is started by opening the page in a web browser, the entire user interface is painted. If a user interface component changes, only the changes are painted. Since Vaadin Release 5, the UIDL communications are currently done using JSON (JavaScript Object Notation), which is a lightweight data interchange format that is especially efficient for interfacing with JavaScript-based AJAX code in the browser. The use of JSON as the interchange format is largely transparent; IT Mill Toolkit version 4 (predecessor of Vaadin released in 2006)

Book of Vaadin

309


User Interface Definition Language (UIDL)

the older versions used an XML-based UIDL representation with the same API. Nevertheless, the UIDL API uses XML concepts such as attributes and elements. Below, we show examples of a Button component in both XML and JSON notation. With XML notation: <button id="PID2" immediate="true" caption="My Button" focusid="1"> <boolean id="v1" name="state" value="false"></boolean> </button>

With JSON notation: ["button", {"id": "PID2", "immediate":true, "caption": "My Button", "focusid":1, "v":{"state":false} } ]

Components are identified with a PID or paintable identifier in the id attribute. Each component instance has its individual PID, which is usually an automatically generated string, but can be set manually with setDebugId() method. Section A.2, “JSON Rendering” gives further details on JSON. For more information about handling UIDL messages in the client-side components, see Chapter 10, Developing Custom Components. You can track and debug UIDL communications easily with the Firebug extension for Mozilla Firefox, as illustrated in Section A.2, “JSON Rendering” below.

A.1. API for Painting Components Serialization or "painting" of user interface components from server to the client-side engine running in the browser is done through the PaintTarget interface. In Vaadin Release 5, the only implementation of the interface is the JsonPaintTarget, detailed in Section A.2, “JSON Rendering” below. The abstract AbstractComponent class allows easy painting of user interface components by managing many basic tasks, such as attributes common for all components. Components that inherit the class need to implement the abstract getTag() method that returns the UIDL tag of the component. For example, the implementation for the Button component is as follows: public String getTag() { return "button"; }

AbstractComponent implements the paint() method of the Paintable interface to handle basic tasks in painting, and provides paintContent() method for components to paint their special contents.The method gets the PaintTarget interface as its parameter.The method should call the default implementation to paint any common attributes. /* Paint (serialize) the component for the client. */ public void paintContent(PaintTarget target) throws PaintException { // Superclass writes any common attributes in // the paint target.

310

API for Painting Components


User Interface Definition Language (UIDL)

super.paintContent(target); // Set any values as variables of the paint target. target.addVariable(this, "colorname", getColor()); }

Serialized data can be attributes or variables, serialized with the addAttribute() and addVariable() methods, respectively. You must always serialize the attributes first and the variables only after that. The API provides a number of variations of the methods for serializing different basic data types. The methods support the native Java data types and strings of the String class. addVariable() also supports vectors of strings. Contained components are serialized by calling the paint() method of a sub-component, which will call the paintContent() for the sub-component, allowing the serialization of user interfaces recursively. The paint() method is declared in the server-side Paintable interface and implemented in the abstract base classes, AbstractComponent and AbstractComponentContainer (for layouts). Layout components have to serialize the essential attributes and variables they need, but not the contained components. The AbstractComponentContainer and AbstractLayout baseclasses manage the recursive painting of all the contained components in layouts. The AbstractField provides an even higher-level base class for user interface components. The field components hold a value or a property, and implement the Property interface to access this property. For example the property of a Button is a Boolean value. public void paintContent(PaintTarget target) throws PaintException { super.paintContent(target); // Serialize the switchMode as an attribute if (isSwitchMode()) target.addAttribute("type", "switch"); // Get the state of the Button safely boolean state; try { state = ((Boolean) getValue()).booleanValue(); } catch (NullPointerException e) { state = false; } target.addVariable(this, "state", state); }

A.2. JSON Rendering Vaadin 5 uses JSON, a lightweight data-interchange format, to communicate UI rendering with the browser, because it is very fast to parse compared to XML. JSON messages are essentially JavaScript statements that can be directly evaluated by the browser. The client-side engine of Vaadin parses and evaluates the UIDL messages with the JSON library that comes with the Google Web Toolkit. Section 3.2.3, â&#x20AC;&#x153;JSONâ&#x20AC;? gave a general introduction to JSON as part of the architecture of Vaadin. In this section, we look into the technical details of the format. The technical details of the JSON messages are useful mainly for debugging purposes, for example using the Firebug plugin for Mozilla Firefox.

JSON Rendering

311


User Interface Definition Language (UIDL)

To view a UIDL message, open the Firebug panel in Firefox, select Net tab, select a "POST UIDL" request, open the Response tab, and click Load Response. This displays the entire UIDL message, as shown in Figure A.1, â&#x20AC;&#x153;Debugging UIDL Messages with Firebugâ&#x20AC;? below.

Figure A.1. Debugging UIDL Messages with Firebug

JSON messages are represented as nested lists and associative arrays (objects with named properties) in JavaScript syntax. At the top level, we can find an associative array with the following fields: changes

Changes to the UI caused by the request.

meta

Meta-information regarding the response and the application state.

resources

Information about application resources.

locales

Locale-specific data for locale-dependent components, such as names of months and weekdays.

The "changes" field contains the actual UI changes as a list of components. Components that can contain other components are represented in a recursive list structure. A component is represented as a list that first contains the UIDL tag of the component, which identifies its class, followed by data fields.The basic representation of component data as attributes and variables is defined in the base classes of the framework. Attributes are represented as an associative array and variables as a separate associative array inside the special "v" attribute. For example, a Button component is communicated with a JSON representation such as the following: ["button", {"id": "PID5", "immediate": true, "caption": "7", "v":{"state":false}} ]

312

JSON Rendering


User Interface Definition Language (UIDL)

A component can give its data also in additional fields in the list instead of the attributes or variables, as is done for the Label component: ["label", {"id": "PID4", "width": "100.0%"}, "Some text here"]

The meta-information field can contain certain types of information, which are not displayed in the UI, but used by the client-side engine. The repaintAll parameter tells that the changes include the entire window contents, not just partial changes. Other data includes redirection details for expired sessions.

JSON Rendering

313


314


Appendix B

Songs of Vaadin Vaadin is a mythological creature in Finnish folklore, the goddess and divine ancestor of the mountain reindeer. It appears frequently in the poetic mythos, often as the trustworthy steed of either Seppo Ilmarinen or Väinämöinen, the two divine hero figures. In many of the stories, it is referred to as Steed of Seppo or Seponratsu in Finnish. An artifact itself, according to most accounts, Vaadin helped Seppo Ilmarinen in his quests to find the knowledge necessary to forge magical artefacts, such as Sampo. Some of the Vaadin poems were collected by Elias Lönnrot, but he left them out of Kalevala, the Finnish epic poem, as they were somewhat detached from the main theme and would have created inconsistencies with the poems included in the epos. Lönnrot edited Kalevala heavily and it represents a selection from a much larger and more diverse body of collected poems. Many of the accounts regarding Vaadin were sung by shamans, and still are. A shamanistic tradition, centered on the tales of Seppo and Vaadin, still lives in South-Western Finland, around the city of Turku. Some research in the folklore suggests that the origin of Vaadin is as a shamanistic animal spirit used during trance for voyaging to Tuonela, the Land of Dead, with its mechanical construction reflecting the shamanistic tools used for guiding the trance. While the shamanistic interpretation of the origins is disputed by a majority of the research community in a maximalist sense, it is considered a potentially important component in the collection of traditions that preserve the folklore. Origin or birth poems, synnyt in Finnish, provide the most distinct accounts of mythological artefacts in the Finnish folklore, as origin poems or songs were central in the traditional magical practices. Vaadin is no exception and its origin poems are numerous. In many of the versions, Vaadin was created in a mill, for which Seppo had built the millstone. After many a year, grinding the sacred acorns of the Great Oak (a version of the World Tree in Finnish mythology), the millstone had become saturated with the magical juices of the acorns. Seppo found that the stone could be used to make tools. He cut it in many peaces and built a toolkit suitable for fashioning spider web into any imaginable shape. When Seppo started making Sampo, he needed a steed that would

Book of Vaadin

315


Songs of Vaadin

help him find the precious components and the knowledge he required. The magical tools became the skeleton of Vaadin. "Lost, his mind was, gone, was his understanding, ran away, were his memories, in the vast land of hills of stone. Make a steed he had to, forge bone out of stone, flesh out of moss, and skin of bark of the birch. The length of his hammer, he put as the spine and the hip, bellows as the lungs, tongs as the legs, paired. So woke Vaadin from the first slumber, lichen did Seppo give her for eating, mead did he give her for drinking, then mounted her for the journey." Other versions associate the creation with Väinämöinen instead of Seppo Ilmarinen, and give different accounts for the materials. This ambiguity can be largely explained through the frequent cooperation between Väinämöinen and Seppo in the mythos. Nevertheless, the identity of the steed or steeds is largely implicit in the myths and, because of the differences in the origin myths, can not be unambiquously associated with a unique identity. The theme of animal ancestor gods is common in the Finnish myth, as we can see in the widespread worship of Tapio, the lord of the bear and the forest. With respect to Vaadin, the identification of the animal is not completely clear. The Finnish word vaadin refers specifically to an adult female of the semi-domesticated mountain reindeer, which lives in the Northern Finland in Lapland as well as in the Northern Sweden and Norway. On the other hand, the Finnish folklore represented in Kalevala and other collections has been collected from Southern Finland, where the mountain reindeer does not exist. Nevertheless, Southern Finnish folklore and Kalevala do include many other elements as well that are distinctively from Lapland, so we may assume that the folklore reflects a record of cultural interaction. The distinction between the northern mountain reindeer and the deer species of Southern Finland, the forest reindeer and the elk, is clear in the modern language, but may not have been in old Finnish dialects. For example, peura, reindeer, may have been a generic word for a wild animal, as can be seen in jalopeura, the old Finnish word for lion. The identification is further complicated by the fact that the line of poems included in Kalevala often refers to a horse. This could be due to the use of the word for horse as a generic name for a steed. While a mountain reindeer is not suitable for riding, animal gods are typically portrayed as uncommonly large in mythology, even to the extremes, so the identification fits quite well in the variety of magical mounts. The mythology related to Vaadin, especially as represented in Kalevala, locates some important characters and people in Pohjola, a mythical land in the north from where all evil originates, according to most accounts. For example, Louhi or Pohjolan emäntä, Queen of Pohjola, is the primary antagonist in the Kalevala mythos. Both Seppo Ilmarinen and Väinämöinen make services to Louhi to earn the hand of her daughters for marriage. Vaadin is often mentioned in connection with these services, such as the making of Sampo. On the other hand, as Sampo can be identified with the mill mentioned in creation stories of Vaadin, its identification in the stories becomes unclear.

316


Songs of Vaadin

While beginning its life as an artifact, Vaadin is later represented as an antropomorphic divine being. This is in contrast with the Bride of Gold, another creation of Seppo, which failed to become a fully living and thinking being. Finding magical ways around fundamental problems in life are central in Kalevala. In some areas, magical solutions are morally acceptable, while in others they are not and the successes and failures in the mythos reflect this ethic. Research in the folklore regarding the Bride of Gold myth has provided support for a theory that creating a wife would go against very fundamental social rules of courting and mating, paralleling the disapproval of "playing god" in acts involving life and death (though "cheating death" is usually considered a positive act). The main motivation of the protagonists in Kalevala is courting young daughters, which always ends in failure, usually for similar reasons. Animals, such as Vaadin, are outside the social context and considered to belong in the same category with tools and machines. The Vaadin myths present a noteworthy example of this categorization of animals and tools in the same category at an archetypal level. The Vaadin myths parallel the Sleipnir myths in the Scandinavian mythology. This connection is especially visible for the connection of Väinämöinen with Odin, who used Sleipnir in his journeys. The use of tongs for the legs of Vaadin actually suggests eight legs, which is the distinguishing attribute of Sleipnir. While Sleipnir is almost universally depicted as a horse, the exact identification of the steed may have changed during the transmission between the cultures. The Bridle of Vaadin is a special artifact itself. There is no headstall, but only the rein, detached from the creature, kept in the hand of the rider. The rein is a chain or set of "gadgets" used for controlling the creature. The rein was built of web with special tools, with Seppo wearing magnifying goggles to work out the small details. The significance and cultural influence of Vaadin can be seen in its identification with a constellation in the traditional Finnish constellation system. The famous French astronomer Pierre Charles Le Monnier (1715-99), who visited Lapland, introduced the constellation to international star charts with the name Tarandus vel Rangifer. The constellation was present in many star charts of the time, perhaps most notably in the Uranographia published in 1801 by Johann Elert Bode, as shown in Figure B.1, “Constellation of Tarandus vel Rangifer in Bode's Uranographia (1801)”. It was later removed in the unification of the constellation system towards the Greek mythology.

317


Songs of Vaadin

Figure B.1. Constellation of Tarandus vel Rangifer in Bode's Uranographia (1801)

318


Index A AbstractComponent, 66, 69 paintContent(), 218 AbstractComponentContainer, 66 AbstractField, 66 paintContent(), 218 AJAX, 5, 37, 38

C caption property, 72 Client-Side Engine, 36, 39, 218 Color Picker, 219 compatibility, 199 Component, 66 Component interface, 68, 69 caption, 72 description, 73 enabled, 74 icon, 75 locale, 75 read-only, 78 style name, 78 visible, 79 cookies, 296-297 CSS, 37, 38, 193-205, 218 compatibility, 199 introduction, 195-200

D Data Model, 37 DefaultWidgetSet, 218 description property, 73 DOM, 38 Drag and Drop, 297-305 Accept Criteria, 300-303

E Eclipse widget development, 220-224 enabled property, 74 events, 37 executeJavaScript(), 275

F Field, 69-72

G getLocale(), 76 Google Web Toolkit, 2, 5, 10, 36, 37, 38, 39, 146, 218, 224, 309 Development Mode, 21 GWT Compiler, 218, 219 GWT Module Descriptor, 218 importing, 240 themeing, 196 widgets, 217-252 GWT Module Descriptor, 218

H HTML templates, 37 HTTP, 36 HttpServletRequest, 295 HttpServletRequestListener, 294-297 HttpServletResponse, 295

I icon property, 75 interfaces, 67 IT Mill Toolkit, 5, 37

J Java, 39, 218 JavaDoc, 67 JavaScript, 2, 38, 218 executeJavaScript(), 275 print(), 275-276 JSON, 36, 37

L layout, 67 Layout, 67 Liferay display descriptor, 282-283 plugin properties, 283-284 portlet descriptor, 282 liferay-display.xml, 282-283 liferay-plugin-package.xml, 283-284 locale property in Component, 75

P Paintable, 68, 218 PDF, 276 portal integration, 276-291 print(), 275-276 printing, 275-276

Book of Vaadin

319


Index

R read-only property, 78

S Sampler, 67 Serializable, 68 server-side component, 218 Sizeable interface, 80 SQL, 38 style name property, 78

T Terminal Adapter, 37 theme, 37, 193-205 tooltips, 73

U UIDL, 36, 37, 218, 309

V VariableOwner, 68 visible property, 79

W widget, definition, 218 widgets, 217 widget set, 218 Window, 67 closing, 276

X XHTML, 38 XML, 38 XMLHttpRequest, 38

320